Information search method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of Internet, in particular to an information search method and device, electronic equipment and a storage medium, and aims to improve the information search efficiency. The method comprises the steps that in response to a search operation triggered based on target search information, first search suggestion information is presented in an information search interface, and the first search suggestion information is summary information of a current search result presented by determining that the current search operation of a target object is related to a historical search operation; in response to a viewing operation triggered based on the first search suggestion information, a search result interface corresponding to the target search information is presented, and a current search result in the interface is generated based on the target search information and context corpus information of the target object. The context corpus information comprises at least one of the following information: search information of each search operation of the target object, content browsed by the target object in each search result, and content that the target object implements a feedback behavior. According to the method, the information search efficiency is effectively improved.

Description

Background Art

[0002] With the popularization of computers and the development of the Internet, people use the network more and more frequently. The computer network has gradually become an essential tool in people's daily lives. The network search service, because it can provide information and data in all aspects for the object, has been widely used in people's daily lives and brought great convenience to people.

[0003] In traditional network searches, when an object needs to solve complex search problems, such as formulating a travel itinerary, the object needs to break down the complex search problem into individual search terms or phrases by itself. For each search term or phrase, the object needs to use a search engine or content platform to conduct a network search separately. After the object conducts multiple searches and browses the content, it then organizes and sorts out the information collected after each search by itself, and finally formulates a travel itinerary that meets its own needs.

[0004] In summary, for some relatively complex search problems, the object needs to conduct multiple cross-platform searches, and the search results need to be summarized by itself. The entire search process is relatively cumbersome, resulting in low search efficiency. Summary of the Invention

[0005] Embodiments of the present application provide an information search method, device, electronic device, and storage medium to improve the information search efficiency of the object.

[0006] An information search method provided by an embodiment of the present application includes:

[0007] In response to a search operation triggered by a target object based on target search information, present first search suggestion information in an information search interface, where the first search suggestion information is a summary information of the current search results determined to meet a preset presentation condition, and the preset presentation condition is that the current search operation of the target object is related to the historical search operations of the target object;

[0008] In response to a view operation triggered based on the first search suggestion information, present a search result interface corresponding to the target search information, where the current search results in the search result interface are generated based on the target search information and the context corpus information of the target object, and the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0009] It should be emphasized that in the specific embodiments of the present application, relevant data related to multiple searches by an object during information search is involved, such as the search information of each search operation of the target object listed above, the content browsed by the target object in each search result, the feedback behavior implemented by the target object, and the content of the feedback behavior implemented by the target object.

[0010] When the above embodiments of the present application are applied to specific products or technologies, the consent or agreement of the object needs to be obtained, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0011] Another information search method provided by an embodiment of the present application includes:

[0012] Obtain the context corpus information of the target object, where the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object;

[0013] After receiving a search request triggered by the target search information, if it is determined that the preset presentation condition is met, input the target search information and the context corpus information into the target language model to generate the current search result and the summary information of the current search result; the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object;

[0014] Feed back the summary information and the current search result to the client, so that after the client determines that the preset presentation condition is met, present the first search suggestion information including the summary information in the information search interface, and present the search result interface including the current search result after responding to the view operation triggered by the first search suggestion information.

[0015] An information search device provided by an embodiment of the present application includes:

[0016] A first response unit, configured to respond to a search operation triggered by a target object based on target search information, and present the first search suggestion information in the information search interface, where the first search suggestion information is the summary information of the current search result presented when it is determined that the preset presentation condition is met, and the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object;

[0017] A second response unit, configured to present a search result interface corresponding to the target search information in response to a viewing operation triggered based on the first search suggestion information, where the current search result in the search result interface is generated based on the target search information and the context corpus information of the target object, and the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0018] Optionally, when the target search information is related to a geographical location, the search result interface further includes a map module; the map module is configured to present: the favorite locations related to the geographical location in the favorite content of the target object, the recommended locations related to the target search information determined based on a preset recommendation rule, and the driving route between the favorite location and the recommended location;

[0019] The current search result in the search result interface includes the basic information of each recommended location and at least one web page result associated with the recommended location.

[0020] Optionally, the second response unit is further configured to:

[0021] Update the search information in the search information input area in the search result interface from the target search information to a combination of the target search information and the number of target favorite contents; the target favorite contents are the contents in the favorite content of the target object that are related to the current search operation; or

[0022] Update the search information in the search information input area in the search result interface from the target search information to the first search suggestion information.

[0023] Optionally, the search result interface further includes additional prompt information that matches the current search result; where the additional prompt information is: prompt information related to at least one of the date information and the location information in the current search process.

[0024] Optionally, the search result interface further includes a smart question-and-answer entry; the device further includes:

[0025] A third response unit, configured to switch from the network search mode to the dialogue search mode in response to a dialogue operation triggered based on the smart question-and-answer entry;

[0026] In response to a question operation for the current search result in the dialogue search mode, present the first question input for the current search result and the first answer information corresponding to the first question.

[0027] Optionally, the search result interface further includes at least one answer operation control; then the device further includes:

[0028] A fourth response unit, configured to, in response to a selection operation on a target answer operation control among the at least one answer operation control, perform an operation corresponding to the target answer operation control on the first answer information, and present a corresponding operation execution result.

[0029] Optionally, the device further includes:

[0030] A fifth response unit, configured to, in response to a content selection operation triggered based on the search result interface, identify the target content selected this time in the search result interface, and the intelligent recognition information corresponding to the target content;

[0031] Wherein, the intelligent recognition information includes at least one of the following:

[0032] The key information included in the target content, the type to which the target content belongs, and the importance level of the target content.

[0033] Optionally, the fifth response unit is further configured to:

[0034] In response to a question operation on the target content, present a corresponding answer interface, where the answer interface includes the target content, a second question input for the target content, and second answer information corresponding to the second question.

[0035] Optionally, the answer interface further includes an intelligent Q&A entry; the fifth response unit is further configured to:

[0036] In response to a dialogue operation triggered based on the intelligent Q&A entry, switch from the network search mode to the dialogue search mode;

[0037] In response to a question operation on the second answer information in the dialogue search mode, present a third question input for the second answer information, and third answer information corresponding to the third question.

[0038] Optionally, the fifth response unit is further configured to:

[0039] Collect the target content, and present a corresponding collection animation in the search result interface; wherein, the collection animation indicates that the target content flies into the collection control in the search result interface.

[0040] Optionally, the device further includes:

[0041] A sixth response unit, configured to present a corresponding collection interface in response to a viewing operation on the collection content of the target object, and in the collection interface, different classification labels are marked with different annotation styles;

[0042] Among them, for the repeatedly occurring classification labels, the number of repeated occurrences is marked at the relevant positions.

[0043] Optionally, the device further includes:

[0044] A summary unit, configured to present at least one second search suggestion message in the information search interface in response to a summary operation triggered by the target object. Each second search suggestion message corresponds to a search topic, and the second search suggestion message is the summary information corresponding to the search topic; the search topic is obtained by dividing multiple search information related to the target object;

[0045] In response to a selection operation on a target search suggestion message among the at least one second search suggestion message, a search result summary interface corresponding to the target search suggestion message is presented. The search result summary interface includes summary content of search results corresponding to each search information under the corresponding search topic.

[0046] Optionally, if the search topic is a travel guide, the summary content is a travel guide including a travel plan corresponding to at least one travel time period, and the collection content referred to by the travel guide is marked;

[0047] If the search topic is knowledge point learning, the summary content includes each knowledge point divided according to the importance degree, and a knowledge example integrating each knowledge point.

[0048] Another information search device provided by an embodiment of the present application includes:

[0049] An information acquisition unit, configured to acquire context corpus information of a target object. The context corpus information includes at least one of the following: search information of each search operation of the target object, content browsed by the target object in each search result, and content of feedback behavior implemented by the target object;

[0050] A result generation unit, configured to, after receiving a search request triggered by target search information, if it is determined that a preset presentation condition is satisfied, input the target search information and the context corpus information into a target language model to generate a summary information of the current search result and the current search result; the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object;

[0051] A feedback unit, configured to feedback the abstract information and the current search result to a client, so that after the client determines that the preset presentation condition is satisfied, a first search suggestion information including the abstract information is presented in an information search interface, and after a viewing operation triggered based on the first search suggestion information is responded to, a search result interface including the current search result is presented.

[0052] Optionally, if the target search information is related to a geographical location, the result generation unit is specifically configured to:

[0053] Determine the collection locations related to the geographical location in the collection content of the target object;

[0054] Input the collection locations into the target language model, use the target language model to learn the feature information of the collection locations, and combine the historical behaviors of the target object to determine at least one recommended location for the target object;

[0055] For each recommended location, by collecting the content related to the recommended location in the network, extract the basic information of the recommended location and at least one associated web page result, and use them as the current search result.

[0056] Optionally, the search result interface further includes additional prompt information matching the current search result; the additional prompt information is: prompt information related to at least one of the date information and the location information in the current search process; the result generation unit is further configured to determine the additional prompt information in the following manner and feedback it to the client through the feedback unit:

[0057] Extract the date information and the location information in the current search process;

[0058] Based on the date information and the location information, generate the prompt information corresponding to each preset prompt template;

[0059] Sort the prompt information corresponding to each prompt template according to their respective preset weights;

[0060] Use the prompt information within the specified order range of the sorting result as the additional prompt information.

[0061] Optionally, if the target search information is related to knowledge point learning and the target language model is a knowledge point recognition model, the result generation unit is specifically configured to:

[0062] Input the knowledge points to be searched corresponding to the target search information into the trained knowledge point recognition model;

[0063] Based on the knowledge point recognition model, if it is determined that the associated knowledge points of the knowledge point to be searched have been searched by the target object, then the knowledge point to be searched, the determined associated knowledge points of the knowledge point to be searched, and the relevant explanations are used as the search result of this time.

[0064] Optionally, the device further includes:

[0065] An intelligent processing unit, configured to determine the selected target content according to the selection range of the target object after the target object makes a content selection for the search result of this time;

[0066] After detecting a voice input event, convert the voice of the question input by the target object for the target content into text;

[0067] Perform semantic analysis and keyword extraction on the text to obtain a text recognition result;

[0068] Input the text recognition result and the target content into the target language model to obtain answer information generated by the target language model for the target content, and feedback it to the client, so that the client presents a corresponding answer interface in response to a question operation for the target content, and the answer interface includes the target content, the question, and the answer information.

[0069] Optionally, the device further includes:

[0070] A summarization unit, configured to respectively determine the search topics corresponding to multiple search information related to the target object after receiving a summarization request triggered by the target object;

[0071] Through the target language model, summarize the search results of different search topics in units of search topics to obtain summary content corresponding to each search topic;

[0072] Score and sort the obtained summary content, and filter out the summary content with a score lower than a preset score threshold;

[0073] Generate second search suggestion information corresponding to the remaining summary content respectively, and feedback the generated second search suggestion information to the client, so that the client presents at least one piece of the second search suggestion information in the information search interface in response to a summarization operation, and presents a search result summary interface corresponding to the target search suggestion information after responding to a selection operation on the target search suggestion information in the at least one piece of second search suggestion information, and the search result summary interface includes the summary content under the corresponding search topic; each piece of second search suggestion information is the abstract information corresponding to the corresponding search topic.

[0074] Optionally, the device further includes:

[0075] A collection unit, configured to classify the collection content of the target object according to a preset dimension, and label the classification tags included in each collection content; the collection content of the target object includes at least one of the following: the content browsed by the target object in each search result, the content on which the target object performs a feedback behavior;

[0076] The summarization unit is specifically configured to:

[0077] Retrieve the target collection content related to the target search information from the collection content according to the classification tags;

[0078] Through the target language model, taking the search topic as a unit, combine the target collection content to summarize the search results of different search topics, and obtain the summary content corresponding to each search topic.

[0079] An electronic device provided by an embodiment of the present application includes a processor and a memory. Among them, the memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of any one of the above information search methods.

[0080] An embodiment of the present application provides a computer-readable storage medium, which includes a computer program. When the computer program runs on an electronic device, the computer program is used to make the electronic device execute the steps of any one of the above information search methods.

[0081] An embodiment of the present application provides a computer program product, the computer program product includes a computer program, and the computer program is stored in a computer-readable storage medium; when the processor of the electronic device reads the computer program from the computer-readable storage medium, the processor executes the computer program, so that the electronic device executes the steps of any one of the above information search methods.

[0082] The beneficial effects of this application are as follows:

[0083] The embodiments of the present application provide an information search method, apparatus, electronic device, and storage medium. Since the present application is based on the network search experience, each time an object (such as a user) conducts a search, the search information input by the object, the content browsed by the object in each search result, and the content of the feedback behavior implemented by the object are all statistically analyzed and constructed into the context corpus information of the object. Furthermore, when the object conducts a search based on the target search information, it first analyzes whether the current search operation is related to the previous historical search operations. If it is related, then the context corpus information of the object and the target search information input by the object this time can be combined and input into the target language model, and the ability of the language model can be utilized to provide the object with the current search result after multi-round information processing that conforms to personal preferences.

[0084] Moreover, when it is determined that the current search operation is related to the previous historical search operations, the first search suggestion information is dynamically presented in the information search interface, prompting the object that the current search is related to the previous search and prompting the summary information of the current search result. After it is determined that the object clicks to view the first search suggestion information, the current search result generated after multi-round information processing is presented. In this way, even for some relatively complex search problems, it is not necessary for the object to conduct multiple cross-platform searches by itself, but the answer that meets the object's needs is automatically generated in combination with the object's personalized corpus information, effectively improving the efficiency of the object's information search.

[0085] Other features and advantages of the present application will be described in the subsequent description, and some of them will become obvious from the description or be understood by implementing the present application. The objectives and other advantages of the present application can be achieved and obtained through the structures specifically pointed out in the written description, claims, and drawings. Brief Description of the Drawings

[0086] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation to the present application. In the drawings:

[0087] Figure 1 is an optional schematic diagram of an application scenario in the embodiments of the present application;

[0088] Figure 2 is the implementation flowchart of an information search method in the embodiments of the present application;

[0089] Figure 3 is a schematic diagram of an information search interface in the embodiments of the present application;

[0090] Figure 4 is a schematic diagram of a search result interface in the embodiments of the present application;

[0091] Figure 5 Schematic diagram of a target content and intelligent recognition information in an embodiment of the present application;

[0092] Figure 6 Animation schematic diagram of collecting target content in an embodiment of the present application;

[0093] Figure 7 Schematic diagram of a first search suggestion information in an embodiment of the present application;

[0094] Figure 8 Schematic diagram of an AI result list and a map module in an embodiment of the present application;

[0095] Figure 9 Schematic diagram of a search result in an embodiment of the present application;

[0096] Figure 10 Schematic diagram of a first additional prompt information in an embodiment of the present application;

[0097] Figure 11 Schematic diagram of a second additional prompt information in an embodiment of the present application;

[0098] Figure 12 Schematic diagram of a third additional prompt information in an embodiment of the present application;

[0099] Figure 13 Schematic diagram of a fourth additional prompt information in an embodiment of the present application;

[0100] Figure 14 Schematic diagram of a first conversational search mode in an embodiment of the present application;

[0101] Figure 15 Schematic diagram of a second conversational search mode in an embodiment of the present application;

[0102] Figure 16 Schematic diagram of a third conversational search mode in an embodiment of the present application;

[0103] Figure 17 Schematic diagram of another information search interface in an embodiment of the present application;

[0104] Figure 18 Schematic diagram of another target content and intelligent recognition information in an embodiment of the present application;

[0105] Figure 19 Schematic diagram of yet another target content and intelligent recognition information in an embodiment of the present application;

[0106] Figure 20 Schematic diagram of a question-asking process in an embodiment of the present application;

[0107] Figure 21 It is a schematic diagram of an answer interface in an embodiment of the present application;

[0108] Figure 22 It is a schematic diagram of another first search suggestion information in an embodiment of the present application;

[0109] Figure 23 It is a schematic diagram of yet another first search suggestion information in an embodiment of the present application;

[0110] Figure 24 It is a schematic diagram of a content summarization process in an embodiment of the present application;

[0111] Figure 25 It is a schematic diagram of another content summarization process in an embodiment of the present application;

[0112] Figure 26 It is a flowchart of the implementation of another information search method provided by an embodiment of the present application;

[0113] Figure 27 It is a schematic diagram of the process of implementing AI multi-round search in network search in an embodiment of the present application;

[0114] Figure 28 It is a schematic diagram of the composition structure of an information search device in an embodiment of the present application;

[0115] Figure 29 It is a schematic diagram of the composition structure of another information search device in an embodiment of the present application;

[0116] Figure 30 It is a schematic diagram of the hardware composition structure of an electronic device applying an embodiment of the present application;

[0117] Figure 31 It is a schematic diagram of the hardware composition structure of another electronic device applying an embodiment of the present application. Detailed implementation manners

[0118] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions of the present application in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, rather than all, of the embodiments of the technical solutions of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments recorded in this application document without creative efforts belong to the scope protected by the technical solutions of the present application.

[0119] The following introduces some concepts involved in the embodiments of the present application.

[0120] Context corpus information: This is a new corpus proposed in this application, which can be used as the context when an object is searched. By integrating the target search information of the object's current search and the context corpus information, the final search result is generated. In the embodiments of this application, for a target object, the context corpus information of the target object includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object. Among them, the feedback behavior refers to some interactive operation behaviors generated when the target object browses the content, such as actively collecting, clicking like, giving a thumbs up, forwarding, sharing, commenting, etc.

[0121] Search suggestion information: This is a type of sug information dynamically presented when an object conducts an online search. According to the change of the query input by the object, the sug generated by the AI each time also changes. In the embodiments of this application, the search suggestion information can be divided into first search suggestion information and second search suggestion information. Among them, the first search suggestion information is the summary information of the current search result determined to meet the preset presentation conditions during a search. The second search suggestion information is the summary information corresponding to the search theme during summarization.

[0122] Additional prompt information: Prompt information related to at least one of the date information and location information during the current search process. In the embodiments of this application, the additional prompt information includes but is not limited to at least one of the following: weather information related to the date, clothing information for travel related to the date, festival activity information related to the date, travel restriction information related to the date, play restriction information related to the location, and welfare information related to the location.

[0123] Intelligent recognition information: Refers to the information obtained by intelligently recognizing the target content selected by the target object, including but not limited to at least one of the following: the key information contained in the target content, the type to which the target content belongs, and the importance level of the target content.

[0124] The embodiments of this application relate to artificial intelligence (AI) and machine learning technologies, and are designed based on natural language processing (NLP) and machine learning (ML) in artificial intelligence.

[0125] AI uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, including theories, methods, technologies, and application systems that can perceive the environment, acquire knowledge, and use knowledge to achieve the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a way similar to human intelligence. Artificial intelligence also studies the design principles and implementation methods of various intelligent machines, enabling machines to have the functions of perception, reasoning, and decision-making.

[0126] Artificial intelligence technology is an interdisciplinary subject with a wide range of fields, including both hardware-level and software-level technologies. The basic technologies of artificial intelligence generally include sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, pre-trained model technology, operation / interaction systems, mechatronics, etc. Among them, the pre-trained model, also known as the large model or the foundation model, can be widely applied to downstream tasks in various directions of artificial intelligence after fine-tuning. The software technology of artificial intelligence mainly includes several major directions such as computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0127] NLP is an important direction in the fields of computer science and artificial intelligence. It studies various theories and methods that can achieve effective communication between humans and computers in natural language. Natural language processing involves natural language, that is, the language people use in daily life, and is closely related to linguistic research; at the same time, it involves computer science and mathematics. The pre-trained model, an important technology for model training in the field of artificial intelligence, evolved from the LLM in the field of NLP. After fine-tuning, large language models can be widely applied to downstream tasks. Natural language processing technology usually includes technologies such as text processing, semantic understanding, machine translation, robot question answering, and knowledge graphs.

[0128] LLM refers to a computer model that can process and generate natural language; it represents a major advancement in the field of artificial intelligence and is expected to change the field through the acquired knowledge. LLM can predict the next word or sentence by learning the statistical laws and semantic information of language data. As the input data set and parameter space continue to expand, the capabilities of LLM will also increase accordingly. It is used in various application fields such as robotics, machine learning, machine translation, speech recognition, image processing, etc., so it is called a Multimodal Large Language Model (MLLM).

[0129] The target language model in the embodiments of the present application is trained using machine learning or deep learning techniques, and the model can be an LLM. Machine learning is a multi-disciplinary field that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behaviors to acquire new knowledge or skills and reorganize the existing knowledge structure to continuously improve their own performance. Machine learning is the core of artificial intelligence and the fundamental way to make computers intelligent, and its applications cover all fields of artificial intelligence. Machine learning and deep learning generally include technologies such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and rote learning. The pre-trained model is the latest development result of deep learning, integrating the above technologies.

[0130] After training the target language model based on the above technologies, the target language model can be applied to generate the answers corresponding to the questions, realizing network search and intelligent dialogue.

[0131] In addition, it should be noted that the target language model in the embodiments of the present application can be trained online or offline, and no specific limitation is made here. In this article, offline training is taken as an example for illustration.

[0132] The following briefly introduces the design concept of the present application:

[0133] With the popularization of computers and the development of the Internet, people use the network more and more frequently. The computer network has gradually become an essential tool in people's daily lives. The network search service, because it can provide various information and data for users, has been widely used in people's daily lives and brought great convenience to people.

[0134] As an important entrance to the network search service, the search engine provides the function of returning search results based on the query keywords. Users can search for questions through the search engine. The search engine recalls the content of the entire network through the search request (query) matching, and scores and ranks the recalled content, and presents it to users in the order of priority. However, when users encounter relatively complex search problems, traditional network searches cannot provide progressive in-depth answers based on the context of multiple user questions. Users need to conduct multiple cross-platform searches, and then process the information and content after each search by themselves, and finally summarize the results that meet their own needs. The entire search process is relatively cumbersome, resulting in low search efficiency.

[0135] However, there are also certain problems with conversational search using large language models. Still taking the travel guide scenario mentioned above as an example, when using conversational search with a large language model and the object asks about the travel itinerary guide for a certain place, the large language model can directly give the object a final complete answer according to its own algorithm. However, the places and content involved in the answer cannot be used to formulate an itinerary based on the object's personal preferences, resulting in the answer being not practical and still requiring the object to summarize it by themselves. In this way, the whole process is rather cumbersome.

[0136] In summary, the current network search or conversational search cannot generate more personalized search results according to the object's own needs, preferences, etc., resulting in the object having to conduct multiple searches and summarize by themselves, with relatively low search efficiency.

[0137] In view of this, the embodiments of the present application propose an information search method, device, electronic device, and storage medium. Since the present application is based on the network search experience, each time the object conducts a search, the search information input by the object, the content browsed by the object in each search result, and the content of the object's implementation of feedback behaviors are all counted and constructed into the context corpus information of the object. Furthermore, when the object conducts a search based on the target search information, first analyze whether the current search operation is relevant to the previous historical search operations. If it is relevant, then the context corpus information of the object and the target search information input by the object this time can be combined and input into the target language model, and the ability of the language model can be utilized to provide the object with the current search result after multi-round information processing that meets the personal preferences.

[0138] Moreover, when it is determined that the current search operation is relevant to the previous historical search operations, the first search suggestion information is dynamically presented in the information search interface, prompting the object that the current search is relevant to the previous search and prompting the summary information of the current search result, so that after it is determined that the object clicks to view the first search suggestion information, the current search result generated after multi-round information processing is presented. In this way, even for some relatively complex search problems, it is not necessary for the object to conduct multiple cross-platform searches by themselves, but rather automatically generate answers that meet the object's needs in combination with the object's personalized corpus information, effectively improving the efficiency of the object's information search.

[0139] The following describes the preferred embodiments of the present application with reference to the accompanying drawings of the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application and are not used to limit the present application. And without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0140] As Figure 1 shown, it is a schematic diagram of the application scenario of the embodiments of the present application. The application scenario diagram includes two terminal devices 110 and a server 120.

[0141] In the embodiments of the present application, the terminal device 110 includes, but is not limited to, devices such as mobile phones, tablet computers, laptop computers, desktop computers, e-book readers, intelligent voice interaction devices, intelligent home appliances, vehicle-mounted terminals, etc.; a client related to information search can be installed on the terminal device, and the client can be software (such as a browser), or a web page, a small program, etc. The server 120 is a background server corresponding to the software, web page, small program, etc., or a server specifically used for information search, and the present application does not make specific limitations. The server 120 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), and big data and artificial intelligence platforms.

[0142] It should be noted that the information search method in each embodiment of the present application can be executed by an electronic device, and the electronic device can be the terminal device 110 or the server 120, that is, the method can be executed independently by the terminal device 110 or the server 120, or jointly executed by the terminal device 110 and the server 120. For example, when jointly executed by the terminal device 110 and the server 120, for example, a client related to information search, such as a browser, is installed on the terminal device 110, and the target object can search in the browser. Each time the target object searches, the server 120 can count the search information input this time, the content browsed by the target object in the search results this time, and the content of the feedback actions (such as liking, collecting, favoring, forwarding, sharing, etc.) implemented by the target object, and construct it into the context corpus information of this object.

[0143] Furthermore, when the target object searches based on the target search information, the browser responds to the search operation triggered by the target object based on the target search information, and sends a search request to the server 120 through the terminal device 110. After receiving the search request triggered by the target search information, the server 120 analyzes whether the current search operation is related to the previous historical search operations. If it is determined to be related (that is, the preset presentation condition is met), the target search information and the context corpus information are input into the target language model to generate the current search result and the summary information of the current search result, and are fed back to the terminal device 110. The terminal device 110 presents the first search suggestion information including the above summary information in the information search interface through the browser; then, after the browser responds to the viewing operation triggered by the first search suggestion information, it can present the search result interface corresponding to the target search information, and the search result interface includes the above current search result.

[0144] In an alternative embodiment, the terminal device 110 and the server 120 can communicate through a communication network.

[0145] In an alternative embodiment, the communication network is a wired network or a wireless network.

[0146] It should be noted that Figure 1 The above is only an example. In fact, the number of terminal devices and servers is not limited and is not specifically defined in the embodiments of the present application.

[0147] In the embodiments of the present application, when the number of servers is multiple, the multiple servers can form a blockchain, and the servers are nodes on the blockchain; for example, in the information search method disclosed in the embodiments of the present application, the information search-related data involved can be stored on the blockchain, such as context corpus information, target search information, target content, first question, first answer information, second question, second answer information, third question, third answer information, etc.

[0148] In addition, the embodiments of the present application can be applied to various scenarios, including but not limited to scenarios such as cloud technology, artificial intelligence, intelligent transportation, and assisted driving.

[0149] It should be emphasized that in the specific implementation of the present application, regarding the relevant data of multiple searches of an object during information search, when the above embodiments of the present application are applied to specific products or technologies, the consent or approval of the object needs to be obtained, and the collection, use, and processing of the relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0150] Next, in combination with the above-described application scenarios, the information search method provided by the exemplary embodiments of the present application will be described with reference to the accompanying drawings. It should be noted that the above application scenarios are only shown for the convenience of understanding the spirit and principle of the present application, and the embodiments of the present application are not limited in this regard.

[0151] Refer to Figure 2 As shown, it is a flowchart of the implementation of an information search method provided by the embodiments of the present application. Taking the client (such as a browser) as the execution subject as an example, the specific implementation process of the method is as follows S21 - S22:

[0152] S21: The client responds to the search operation triggered by the target object based on the target search information, and presents first search suggestion information in the information search interface. The first search suggestion information is the summary information of the search results presented this time determined to meet the preset presentation conditions, and the preset presentation conditions are: the current search operation of the target object is related to the historical search operations of the target object.

[0153] Among them, the first search suggestion information refers to the summary information of the search results during this search. This information is presented in the information search interface only when this search meets the preset presentation conditions.

[0154] Specifically, every time an object conducts a network search, the current search operation of the object can be compared with the previous historical search operations of the object to analyze whether there have been relevant search behaviors before.

[0155] Taking the target object as an example, the search information input by the target object during this search can be recorded as the target search information. Specifically, the target search information can be compared with the search information, search results, etc. corresponding to the previous historical search operations of the target object to analyze whether they are relevant.

[0156] Among them, relevant means that there is a certain commonality between two search operations. This commonality can specifically refer to that the search information contains the same or similar keywords (such as names of people, places, organizations, etc.), the semantics of the search information are similar, the semantics of the search information are opposite (which can also be understood as a kind of commonality, semantic relevance), etc., and will not be elaborated one by one here.

[0157] When it is determined that the current search operation of the target object is relevant to the historical search operations of the target object, that is, it is determined that the preset presentation conditions are met, the first search suggestion information can be dynamically presented in the information search interface.

[0158] In the embodiments of the present application, according to the change of the query (i.e., search information) input by the object, the sug generated by the AI each time also changes. And the AI sug appears during the object's search process, and can prompt the object with the answer given by the AI at the first time when the object inputs, providing the object with a quick preview of the AI search results combined with the context. By browsing the AI sug, the object can judge whether it is necessary to view the content generated by the AI combined with the context.

[0159] It should be emphasized here that in the above process, it is necessary to compare and analyze the current search operation of the target object with the historical search operations. In this case, it will involve data related to multiple searches of the target object, and the collection, use, and processing of these data comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0160] And the acquisition of these data requires the permission or consent of the target object. Specifically, the target object can be asked before the analysis whether it is necessary to obtain the historical search operations of the target object for comparison, etc. After the target object agrees, the above analysis process can be further executed; or, the target object can also be prompted that it is necessary to obtain the historical search operations of the target object for comparison, etc.

[0161] S22: In response to a viewing operation triggered by the first search suggestion information, the client presents a search result interface corresponding to the target search information. The search results in the search result interface this time are generated based on the target search information and the context corpus information of the target object. The context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0162] Among them, the search information refers to the query entered by the target object when performing an online search, such as the search term entered by the target object in the search box.

[0163] The feedback behavior refers to some interactive operation behaviors generated when the target object browses the content, such as actively collecting, clicking like, giving a thumbs up, forwarding, sharing, commenting, etc.

[0164] Specifically, in each online search process of the target object, the above-mentioned types of information can be collected, thereby continuously enriching the context corpus information corresponding to the target object for the generation of subsequent personalized search results.

[0165] It should be emphasized that the acquisition of data such as the above feedback behavior has also obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0166] In the embodiment of the present application, the generation of the search results this time can be implemented based on a target language model (such as a large language model). Specifically, each time the target object performs an online search, the current search operation can be compared with the previous historical search operations. When it is determined that the preset presentation conditions are met, the target search information entered by the target object this time and the currently collected context corpus information are input into the target language model together. Using the answer generation ability of the target language model and combining the context corpus information, a search result (i.e., the search result this time) that meets personal preferences and has undergone multi-round information processing is generated at one time.

[0167] In the embodiment of the present application, the object can judge whether to view the content generated by the AI in combination with the context by browsing the AI sug. When the object determines that it needs to view, it can trigger the viewing operation based on the first search suggestion information and view the search results generated by the AI this time.

[0168] Specifically, the object can trigger this viewing operation by clicking on the first search suggestion information, long pressing the first search suggestion information, etc., which will not be elaborated here one by one.

[0169] Based on the network search experience, this application combines a target language model and provides all the search questions the object conducts in the browser, the content browsed in the search results, and the content actively liked and collected by the object, etc., as context corpus information to the target language model. The target language model can combine the customized context corpus information of the object and utilize the capabilities of the target language model to provide the object with search results after multi-round information processing that conform to personal preferences, solving the problem that current network searches can only perform single searches and cannot perform multi-round searches.

[0170] In addition, this application also solves the problem that the target language model cannot provide answers based on the object's personalized corpus. In this application, the active and passive operations of the object regarding the content in the network search are provided to the target language model as the object's personalized corpus, enabling the target language model to provide personalized solutions for the object.

[0171] Meanwhile, this application also solves the problem of the mutual integration of network search and the target language model: when the object uses network search, they can click on AI sug to view the results screened by AI at any time.

[0172] Next, taking the two scenarios of travel and word query as examples, the information search method in the embodiments of this application will be described in detail with reference to the accompanying drawings:

[0173] (1) Travel scenario.

[0174] Next, taking the example where the target object first searches for "Travel guide for City A" and then searches for "Hotels in City A":

[0175] Refer to Figure 3 shown, which is a schematic diagram of an information search interface in the embodiments of this application. Figure 3 It represents that the target object first searches for "Travel guide for City A". Assuming that the object has not searched for travel-related or City A-related content before, that is, there is no historical search operation related to the current search operation and it does not meet the preset presentation conditions. Therefore, the first search suggestion information does not need to be presented in the information search interface at this time.

[0176] After the target object conducts a network search based on the input search information "Travel guide for City A" this time, a common search result interface can be presented. The search results in this search result interface are the content recalled by the search engine based on "Travel guide for City A" from the entire network, and after scoring and sorting the recalled content, they are presented in the order of priority of the sorting.

[0177] As Figure 4 shown, which is a schematic diagram of a search result interface in the embodiments of this application. As Figure 4The left - hand interface lists several search results sorted and recalled in the above - mentioned manner, such as Search Result 1 "Nine Favorite Places to Visit in City A", Search Result 2 "Spent Four Days in City A, Sincere Suggestions", and Search Result 3 "The Most Beautiful Scenic Spots in City A in Autumn". The target object can swipe up the interface to view more search results or click to view a specific search result. If the object selects Search Result 1, it can be redirected to Figure 4 the details page of Search Result 1 shown on the right.

[0178] In the embodiment of the present application, if the target object browses the articles of interest in the search result interface, the content browsed and collected by the target object can be automatically recorded and created into a personalized corpus of the object (i.e., a form of expression of the personalized corpus information in this article).

[0179] It should be noted that in the embodiment of the present application, the collected content of the target object can not only include the content actively clicked and collected by the target object, but also include the content that the target object clicks like, thumbs - up, comments, forwards, shares, etc. In addition, when the target object browses the search results, the target object can also mark and question the partial content in the search results. The embodiment of the present application also supports intelligent collection of this part of the content to enrich the dimension of the collected content of the target object.

[0180] It should be emphasized here that the relevant processing of automatically recording the content browsed and collected by the target object mentioned above has obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0181] Specifically, when the target object browses an article and is interested in a partial content of it, for such content, AI intelligent recognition or AI intelligent collection can be performed. An optional implementation method is as follows:

[0182] The client responds to the content selection operation triggered based on the search result interface, marks the target content selected this time in the search result interface, and the intelligent recognition information corresponding to the target content.

[0183] Among them, the intelligent recognition information includes but is not limited to at least one of the following:

[0184] The key information contained in the target content, the type to which the target content belongs, and the importance level of the target content.

[0185] Among them, the key information contained in the target content can refer to people, places, institutions, special foods, special activities, special events, etc.

[0186] In the embodiments of the present application, for any search operation, the target object can further view the details of the search result interface. For example, when the target object views a specific search result in the search result interface and enters the details page of this search result (which belongs to another type of search result interface in the embodiments of the present application), furthermore, when performing a content selection operation on this details page, the client, in response to the content selection operation of the target object, in addition to marking the selected target content in the search result interface, can further present the intelligent recognition information corresponding to this target content.

[0187] Specifically, the intelligent recognition information can be in the form of pictures, texts, text and pictures, videos, icons, labels, animations, etc., which are not specifically limited herein. Hereinafter, an example will be given taking labels as an example.

[0188] As Figure 5 shown, it is a schematic diagram of a target content and intelligent recognition information in the embodiments of the present application. For example, when the target object is browsing the article "9 Favorite Places to Travel Around City A" in the above-listed search result 1 and is interested in a certain scenic spot and route mentioned therein, after double-clicking on the content area, the paragraph where the object double-clicks will be selected on the page. As Figure 5 the part S501 marked by the gray area in.

[0189] It should be noted here that if the selected range is inaccurate, the target object can manually drag the upper and lower edges of the selection area to expand or shrink it to modify the selected range.

[0190] After the target object has no operation, an identification animation can appear on the page to prompt the object that "AI is identifying the selected content" ( Figure 5 not shown in). After the identification animation ends, information such as the location identified by AI can be marked on the page and presented in the form of labels, such as Figure 5 S502 and S503 in the right interface in. S502 represents the identified location "Palace B", and S503 represents the local special cuisine "D Dish".

[0191] In the embodiments of the present application, for different intelligent recognition information, different styles can also be used for marking. For example, different types of locations are marked with different colors, etc., which will not be elaborated here one by one.

[0192] Optionally, in addition to intelligently recognizing the target content and marking the intelligent recognition information, the target content can also be automatically collected. Specifically, when the target object selects a certain part of the content for collection, when collecting, AI uses the target language model to analyze and label the collected content, and classifies the content that the target object is interested in intelligently when collecting. The classification labels can be used in the retrieval of the object's personalized corpus during AI conversations. There is a relevant introduction on the server side, which will not be elaborated here one by one.

[0193] In addition, a corresponding favorite animation is presented in the search result interface. Among them, an optional favorite animation shows the target content flying into the favorite control in the search result interface.

[0194] In this process, the target content can also gradually shrink, giving the target object a feeling that the target content is getting smaller and smaller, and finally shrinking to a size that can enter the favorite control.

[0195] Such as Figure 6 As shown, it is a schematic diagram of an animation for favoriting target content in an embodiment of the present application. Figure 6 The interface shown on the left represents a frame in the process of the target content S601 flying into the favorite control in the search result interface. In this process, S601 not only includes the target content, but also includes labels recognized by AI, such as B Palace and D Dish. After the target content completely flies into the favorite control, the favorite control is updated from the style shown by S602 to the style shown by S603.

[0196] In an embodiment of the present application, the target object can click on the favorite control to view the favorite folder of the target object. In the favorite folder of the target object, the intelligent recognition information can also be marked in the same or similar manner as above, such as still marked in the form of labels, which is convenient for generating answers later.

[0197] Optionally, the client can also respond to the viewing operation of the favorite content of the target object, present a corresponding favorite interface, and in the favorite interface, different classification labels are marked with different marking styles; among them, for the repeated classification labels, the number of repeated occurrences is marked at the relevant position.

[0198] Specifically, these classification labels can be location labels, time labels, event labels, item labels, etc. For specific details, please refer to the relevant description on the server side and will not be repeated here.

[0199] In the above implementation, a way to favorite local content is provided for the object. Secondly, in the traditional intelligent favorite of content, only the text or picture itself is favored, and the favored content cannot be classified or recognized. However, in the above implementation, the object can see the labels marked by AI for the key content when favoriting, which is convenient for the object to better organize and understand the content during favoriting.

[0200] It should be noted that the above Figure 5 and Figure 6That is, it is the related operation on the search results when the target object conducts the first search. Of course, when the target object conducts any subsequent search, the above operation can be performed on the search results. For example, when the target object conducts subsequent searches related to "City A" and "travel", the above operation can also be performed on the search results. The specific implementation manner is the same as that of the above embodiment, and the repeated parts will not be elaborated.

[0201] Suppose that after the target object conducts the first search for "travel guide of City A", a second search is conducted, and this time "hotels in City A" is searched. It is related to the previous search for "travel guide of City A" and both belong to search operations related to City A.

[0202] Therefore, in the embodiment of the present application, when the target object searches for "hotels in City A", the preset presentation condition is satisfied, and during this search process, AI sug, that is, the first search suggestion information, can be dynamically presented on the information search interface.

[0203] Such as Figure 7 shown, it is a schematic diagram of a kind of first search suggestion information in the embodiment of the present application. During this search process, AI sug can be presented in the Figure 7 shown information search interface, prompting that relevant hotels in City A have been screened for the target object. As Figure 7 shown in S701 and S702 in it, where the AI sug shown in S701 represents the current processing progress, that is, no suitable hotel has been searched yet, but it is in the stage of ongoing analysis; after the analysis is completed, the AI sug shown in S702 can be presented. This AI sug is the first search suggestion information in the embodiment of the present application, prompting the summary information of the search results for the target object. For example, in S702, it is "Based on your preferred location, recommend hotels in City A for you".

[0204] In the embodiment of the present application, the target object can click on the AI sug shown in S702 to enter the search result interface and present the AI result list. As Figure 8 shown, it is a schematic diagram of a kind of AI result list and map module in the embodiment of the present application. Figure 8 In this AI result list, the hotels near the scenic spots, routes, and stores that the target object has collected or followed in previous browsing are screened by AI, which is an AI personalized screening result that meets the needs of the target object.

[0205] It should be noted that during the information search process, in addition to presenting the first search suggestion information, search suggestion information indicating the generation progress of the search results (which can also be denoted as the third search suggestion information) such as that shown in S701 can also be presented.

[0206] In addition, it should be noted that during multiple network searches by the target object, the target language model can cross-screen the previous search terms of the object, screen the results in the current search result interface for the object, and display them in a suitable form according to the types of the screened content. Or, compare multiple search terms with each other and give the results after comparison.

[0207] For example, if the target search information is related to a geographical location, a map module can be further set in the search result interface; this map module is used to present: the collection locations related to the geographical location in the collection content of the target object, the recommended locations related to the target search information determined based on preset recommendation rules, and the driving routes between the collection locations and the recommended locations; furthermore, the current search results in the search result interface include the basic information of each recommended location and at least one web page result associated with the recommended location.

[0208] Still taking Figure 8 the example shown, where the S801 part is an example of a map module in the embodiment of the present application. In Figure 8 the example shown, the content screened by AI is related to the geographical location "City A", and the map module shown in S801 is added. The locations collected by the target object and the hotel locations recommended by AI are presented on the map module at the same time, and the route between the two is marked. As shown in S801, the collection locations of the target object are Scenic Spot x1, Scenic Spot x2, Scenic Spot x3, and Palace B, and the recommended locations by AI are Holiday Hotel P1, Xixia Xiaozhu P2, and Holiday Hotel P3.

[0209] In addition, the map module also marks the routes between the collection locations and the recommended locations. Specifically, according to the distances between these locations, a feasible route can be marked between the locations within a certain range (relatively close). For example, there is a recommended route between Holiday Hotel P1 -> Scenic Spot x2 -> Palace B, a recommended route between Xixia Xiaozhu P2 -> Scenic Spot x1, and a recommended route between Holiday Hotel P3 -> Scenic Spot x3. These three recommended routes are shown as the dotted lines in S801.

[0210] Among them, there may be duplicate locations between the collection locations of the target object and the recommended locations by AI, which are not specifically limited in this article.

[0211] It should be noted that the optional range of hotel prices is also presented in the map module shown in S801. For example, the prices of the three recommended hotels in S801 are between 0 and 1000 yuan. This price range can be obtained through statistical analysis based on the historical reservation records of the target object, etc., or it can be a default price range, which is not specifically limited in this article.

[0212] Furthermore, part S802 presents a list of the above-mentioned hotels selected by AI, and the presentation of the hotel list includes two elements: one is the hotel information (i.e., basic information of the recommended place), and the other is the web page results selected by AI that mention the hotel (i.e., at least one web page result associated with the recommended place).

[0213] Among them, one hotel can be followed by one or more web page results to help the target object determine whether he is interested in the current hotel.

[0214] It should be emphasized here that the above-mentioned embodiments require obtaining the target object's favorite locations, some historical search records of the target object, the target object's historical reservation records for hotels and other data. The acquisition of these data is subject to the object's permission or consent, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0215] The following example takes each hotel as an example. For details of S802, please refer to Figure 9 ,like Figure 9 As shown, it is a schematic diagram of a search result in an embodiment of the present application. Figure 9 The hotel list shown introduces P1 Resort Hotel, P2 Xixia Cottage and P3 Resort Hotel in turn, where the basic information of the hotel is the name, price, location, some architectural pictures and other information of the hotel. In addition, each hotel also presents at least one web page result. For example, there is one guide that mentions P1 Resort Hotel, and the content comes from platform a. The content of the web page result is shown in part S901. Similarly, there is one guide that mentions P2 Xixia Cottage, and the content comes from platform b. The content of the web page result is shown in part S902; there is one guide that mentions P3 Resort Hotel, and the content comes from platform a. The content of the web page result is shown in part S903.

[0216] In the above implementation, the subject can filter content more efficiently, without filtering in the entire network results, but filtering in combination with existing preferred content. In addition, AI intelligent filtering can use different components according to the different forms of the subject's input content, such as the map listed above. Compared with network search and target language model dialogue, it can display content more intuitively.

[0217] Optionally, after the target object triggers a viewing operation based on the first search suggestion information, the search result interface corresponding to the target search information can be presented. At the same time, considering that the current search results in this search result interface are generated based on the target search information and the context corpus information, which is different from the way of directly matching and recalling the entire network content based on the target search information in traditional web search. Therefore, the search information in the search information input area of the search result interface can also be updated to more intuitively show the difference between AI intelligent screening and traditional web search to the target object.

[0218] An optional implementation is: update the search information in the search information input area of the search result interface from the target search information to the combination of the target search information and the number of target favorite contents; the target favorite contents are the contents related to the current search operation in the favorite contents of the target object.

[0219] Still taking Figure 8 as an example, where Figure 8 the search information input area in is the search box shown in S803. The search information in this search box was originally supposed to be the target search information, such as Figure 8 shown on the left interface, "Hotels in City A", but in S803, it is updated to be displayed in two parts. One part is the target search information "Hotels in City A", and the other part is "☆4 favorite locations".

[0220] In the above implementation, on the result page of AI screening, the query in the search box is updated to "the query input by the target object" and "※ favorites", prompting the target object that the current AI screening result is generated by combining the query input by the object this time and the locations previously favorited by the object, which is more in line with the preferences of the target object and more meets the personalized needs of the target object.

[0221] Another optional implementation is: update the search information in the search information input area of the search result interface from the target search information to the first search suggestion information.

[0222] For example, directly update the search information in S803 to "Based on your preferred locations, we recommend hotels in City A for you".

[0223] In this way, it can more directly prompt the target object that the current AI screening result is generated by combining the query input by the object this time and the object's preferences, which is more in line with the preferences of the target object and more meets the personalized needs of the target object.

[0224] It should be noted that the above two methods for updating search information are only simple examples. In addition, any method that can prompt that the current AI screening result of the object is generated by combining the query input by the object this time and the preferences of the object is applicable to the embodiments of this application, and will not be elaborated one by one here.

[0225] In the embodiments of this application, intelligent reminders can also be made for the results of AI screening. That is, in addition to screening content according to the personalized information provided by the object, additional reminder information can also be matched according to the searched content. An optional implementation method is as follows:

[0226] Additional prompt information that matches the current search results is also presented on the search result interface; wherein, the additional prompt information is: prompt information related to at least one of the date information and location information in the current search process.

[0227] Generally speaking, in the current search process, if date information is involved, additional prompt information related to the date can be further generated; similarly, if location information is involved, additional prompt information related to the location can be further generated; or, when both date information and location information are involved, in addition to the above-mentioned individual generation methods, additional prompt information related to both the date and the location can also be generated.

[0228] The following is a simple example to illustrate the additional prompt information:

[0229] In the embodiments of this application, the additional prompt information includes but is not limited to at least one of the following:

[0230] Weather information related to the date, clothing information for travel related to the date, festival activity information related to the date, travel restriction information related to the date, play restriction information related to the location, welfare information related to the location.

[0231] Specifically, the above additional prompt information can be generated by matching a pre-set prompt template, and these additional prompt information can be in the form of pictures, texts, graphics and texts, videos, icons, labels, animations, etc., which are not specifically limited in this article.

[0232] For example, when the target object provides date information, the prompt templates that can be matched include but are not limited to: the weather corresponding to the date, the clothing for travel corresponding to the date, the major festivals corresponding to the date, the travel restriction information corresponding to the date, etc.

[0233] When the target object does not provide date information, the prompt templates that can be matched include but are not limited to: play restriction information of scenic spots, welfare information of merchants, etc.

[0234] As Figure 10 shown, it is a schematic diagram of the first additional prompt information in the embodiments of this application.Figure 10 In the search result interface shown, before the hotel list in the intelligent answer section, additional prompt messages S1001 and S1002 are presented. For example, if the target search information provided by the object contains the date information "z festival", S1001 is the weather information corresponding to a certain z festival listed in this application, and S1002 is the travel clothing information corresponding to a certain z festival listed in this application.

[0235] Another example is Figure 11 As shown, it is a schematic diagram of the second type of additional prompt message in the embodiment of this application. Figure 11 In the search result interface shown, before the hotel list in the intelligent answer section, additional prompt message S1101 is presented. For example, if the target search information provided by the object contains the date information "z festival", S1101 is the travel restriction information corresponding to a certain z festival listed in this application, and the specific content is "During the z festival, there is traffic control due to the y sports meeting being held. Recommended hotels that are more convenient for walking are provided for you."

[0236] Another example is Figure 12 As shown, it is a schematic diagram of the third type of additional prompt message in the embodiment of this application. Figure 12 In the search result interface shown, before the hotel list in the intelligent answer section, additional prompt message S1201 is presented. For example, if the target search information provided by the object does not contain date information but contains location information "A city", and some locations such as scenic spots and hotels related to A city are further involved in the search process, S1201 is the play restriction information corresponding to a certain scenic spot listed in this application, and the specific content is "For the x4 scenic spot you plan to visit, women are not allowed to enter wearing skirts or off-the-shoulder tops. Remember to prepare appropriate clothes~".

[0237] Another example is Figure 13 As shown, it is a schematic diagram of the fourth type of additional prompt message in the embodiment of this application. Figure 13 In the search result interface shown, before the hotel list in the intelligent answer section, additional prompt message S1301 is presented. For example, if the target search information provided by the object does not contain date information but contains location information "A city", and some locations such as scenic spots and hotels related to A city are further involved in the search process, S1301 is the welfare information corresponding to a certain merchant listed in this application, and the specific content is "The recommended hotel for you has a free shuttle bus that can go to scenic spot c and scenic spot x4~".

[0238] It should be noted that the above-listed several types of additional prompt messages are only simple examples. In addition, other additional prompt messages related to dates and / or locations are applicable to the embodiments of this application and will not be elaborated one by one here.

[0239] In the above embodiments, by presenting additional prompt information, more personalized information is provided for the object, which better meets the object's own needs and reduces the object's search path.

[0240] In addition to the above embodiments, the information search method in the embodiments of the present application also supports switching the AI screening results to the conversational search mode. For example, it is set that the search result interface further includes an intelligent Q&A entry, and the target object can switch the mode based on this intelligent Q&A entry. One optional embodiment is as follows:

[0241] The client responds to the conversation operation triggered based on the intelligent Q&A entry and switches from the network search mode to the conversational search mode; then, the target object can ask questions about the current search results. The client responds to the question operation about the current search results in the conversational search mode and presents the first question input for the current search results and the first answer information corresponding to the first question.

[0242] Specifically, after the target object clicks on AI sug and sees the AI content, they can continue the AI conversation in the form of directly talking to the target language model through the "AI conversation" entry, or they can click on the search box to conduct a network search.

[0243] As Figure 14 shown, it is a schematic diagram of the first conversational search mode in the embodiments of the present application. As Figure 14 shown, the intelligent Q&A entry in the search result interface is the "Continue asking AI" control of S1401. When the target object clicks on S1401 in the hotel list screened by AI to continue the conversation, the network search mode can be switched to the conversational search mode. The target object can directly continue to ask questions based on the current search results. For example, the target object continues to enter a question (which can also be understood as a kind of conversation instruction), such as S1402 "Help me make a 3-day and 2-night travel itinerary for City A with a budget of 4000 yuan". The AI combines the target object's personalized corpus (including favorited content, browsed content, etc.), the content retrieved from the network, and the target language model to give the answer (i.e., the first answer information) corresponding to this question (i.e., the first question).

[0244] As Figure 14 shown in S1403 of

[0245] Specifically, after analyzing all the relevant content favorited by the target object, network retrieval can be performed to search the entire network for content related to traveling in City A. As Figure 15As shown, it is a schematic diagram of the second conversational search mode in the embodiment of the present application. In part S1501, it is prompted that the object is currently "generating a travel guide based on your preferences". Two steps have been completed in sequence before this step, namely "analyzing your collection content" and "searching the whole network for travel in City A". It is prompted that the object has analyzed 32 collection contents related to the current problem, and generates an answer by combining the collected content and the content retrieved from the network.

[0246] As Figure 16 shown, it is a schematic diagram of the third conversational search mode in the embodiment of the present application. Figure 16 Part S1601 shown indicates that the above three steps are all completed, and the specific content of the first answer information generated is displayed after S1601, that is, a 3-day and 2-night travel guide for City A with a budget of 4,000 yuan. As in part S1602, the itinerary for each day and the accommodation location are introduced in detail. For example, on the first day, visit scenic spots x4 -> B Palace -> scenic spot c, and make a plan according to time periods such as morning, afternoon, and evening, listing the specific itinerary plan. Similarly, the itinerary plans for the second and third days are also given. For specific details, please refer to Figure 16 , and will not be elaborated here one by one.

[0247] For the locations from the object's collection involved in the generated travel guide, special styles can also be used for identification (such as five-pointed stars) to prompt the object that the location comes from the object's collection. Specifically, these identifications can be directly marked at the corresponding positions where the location appears in the travel guide. They can also be marked together at a unified and fixed position. As shown in part S1603 of Figure 16 , after the specific itinerary plan, it is uniformly marked that the content reference comes from the marked collection, such as historical block x1, B Palace Museum, and P1 resort hotel.

[0248] In the above implementation, after the object switches to the AI conversation, each time the AI answers, it will combine the results of network search, the object's personalized corpus, and the target language model to answer the question.

[0249] It should be emphasized here that in the above-mentioned embodiments, data such as the network search results and personalized corpus of the target object need to be obtained. The acquisition of these data has obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0250] Optionally, the search result interface may further include at least one answer operation control; the target object may perform related operations on the generated first answer information based on these operation controls. The client responds to the selection operation on the target answer operation control among the at least one answer operation control, performs the operation corresponding to the target answer operation control on the first answer information, and presents the corresponding operation execution result.

[0251] Still taking Figure 16 as an example, as shown in part S1604, several answer operation controls listed in this application are: share link, download PDF, add to favorites. That is to say, for the generated strategy content, the target object can perform operations such as sharing, downloading, and collecting.

[0252] It should be noted that Figure 16 the several answer operation controls listed in

[0253] are only simple examples. In addition, they can also be other controls, such as like, forward, etc., which will not be elaborated here one by one.

[0254] In the above embodiment, the object can also share, download, collect, etc. the answers generated by the AI. While providing a more abundant interaction method for the object, it is more helpful to improve the object's usage stickiness and provide more convenience for the object.

[0254] In summary, when the object needs to solve complex search problems, such as formulating a travel strategy, the object does not need to perform multiple cross-platform searches. This application can provide progressive in-depth answers through the context, and the object does not need to process the information and content after each search by itself. And in the case of formulating a travel strategy, the places and content involved are all formulated according to the object's personal preferences. This application is based on the network search experience and combines the target language model. Every search question the object conducts in the browser, the content browsed in the search results, and the content the object actively likes and collects, etc., are all provided to the target language model as context corpus information. The target language model can combine the object's customized corpus and use the capabilities of the target language model to provide the object with search results that meet personal preferences and have undergone multi-round information processing.

[0255] The above uses the travel scenario as an example to expand and explain the information search method in the embodiments of this application. The following will be expanded and explained with the word query scenario as an example:

[0256] (2) Word query scenario.

[0257] The following takes the target object's first search for the word "artificial" and then searches for the word "naturally" as an example:

[0258] Refer to Figure 17As shown, it is a schematic diagram of another information search interface in an embodiment of the present application. Figure 17 It indicates that the target object has searched for the word "artificial" for the first time. Assuming that the object has not searched for other words before, that is, there is no historical search operation related to this search operation, and the preset presentation conditions are not met. Therefore, there is no need to present the first search suggestion information in the information search interface at this time.

[0259] After the target object searches the Internet based on the search information "artificial" input this time, a normal search result interface may be presented. The search results in the search result interface are the search results of "artificial" obtained by the search engine based on the search information "artificial". For example, Figure 17 As shown in the right interface, the meaning of the word and related phrases, examples, etc. are introduced in detail, and the target object can browse the search results of the word.

[0260] Similar to the above travel scenario, after obtaining the search results, the target object can also browse the search results. If the target object is interested in the partial content of the search results, the target object can perform AI intelligent recognition and AI intelligent collection for such content. An optional implementation is as follows:

[0261] In response to the content selection operation triggered based on the search result interface, the client identifies the target content selected this time and the intelligent identification information corresponding to the target content in the search result interface.

[0262] Among them, the intelligent identification information is specifically referred to the above embodiment, and the repeated parts are not repeated here.

[0263] like Figure 18 As shown, it is a schematic diagram of another target content and intelligent recognition information in an embodiment of the present application. For example, when the target object browses the above search results, it is interested in a phrase "artificial intelligence" mentioned therein. After double-clicking the content area, the phrase of the double-clicked part of the object will be selected on the page.

[0264] After the target object has no operation, a recognition animation can appear on the page to prompt the object that "AI is recognizing the selected content" ( Figure 18 After the recognition animation is finished, the page can identify the type and importance of the target content identified by AI, and present it in the form of a label, such as Figure 18 As shown in S1801, the type of the target content is "phrase" (also meaning a phrase) and the importance is three stars.

[0265] In addition, the target content can be automatically collected and the corresponding collection animation will be displayed, such asFigure 18 S1802 shown in the right interface in [figure reference] is a frame of the favorite animation in the embodiment of the present application, indicating that the target content flies into the favorite control.

[0266] As Figure 19 shown, it is a schematic diagram of another target content and intelligent recognition information in the embodiment of the present application. For example, when the target object browses the above search results and is interested in a certain example sentence "They make up something artificial or untrue. Make something artificial or untrue..." mentioned therein, after double-clicking on the content area, the example sentence where the object double-clicked will be selected on the page.

[0267] After the target object has no operation, an identification animation can appear on the page to prompt the object that "AI is identifying the selected content" ( Figure 19 not shown in [figure reference]). After the identification animation ends, information such as the type and importance level of the target content identified by AI can be marked on the page and presented in the form of labels. As Figure 19 shown by S1901 in [figure reference], the type of this target content is "example sentence" and the importance level is five stars.

[0268] In addition, the target content can be automatically favorited and the corresponding favorite animation can be presented. As Figure 19 shown by S1902 in the right interface in [figure reference], it is a frame of the favorite animation in the embodiment of the present application, indicating that the target content flies into the favorite control.

[0269] It should be emphasized here that the processing process of automatically favoriting the search results browsed by the target object in the above-mentioned embodiments has obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0270] In addition to the above interaction methods, the present application also supports the object to perform multimodal AI search on any content. In any scenario on the browsing result page, after the object long-presses and selects text or a picture, the object can ask questions about the selected content. The target language model will combine the selected content and the object's question to give an answer.

[0271] An optional implementation method is as follows:

[0272] The client responds to the question operation for the target content and presents the corresponding answer interface, which includes the target content, the second question input for the target content, and the second answer information corresponding to the second question.

[0273] In the embodiment of the present application, the target object can also select the target content during the process of browsing the search results and ask questions about the target content through voice input, keyboard input, etc.

[0274] For example, when the target object is learning a certain word or browsing any web page content, if the target object has a question about the content on the page, the target object can directly long-press the page content. After the long-press, voice collection is activated, and the target object can directly speak out the question. After the target object releases the finger, the selected page content and the question described by the target object will be sent to the AI together, and the AI will return the result.

[0275] As Figure 20 shown, it is a schematic diagram of a question-asking process in an embodiment of the present application. Figure 20 That is, it means that during the process of the target object browsing the search results of the word "artificial", the target object has a question about some of the content. The target object can select the content with questions as the target content by double-clicking, long-pressing, etc. Then, by double-clicking, long-pressing, etc., voice collection is activated, and the target object can speak out his question, such as Figure 20 That is, it means that the target object has selected the target content "This artificial beach resort can hold up to 10,000 visitors.", and the target object is prompted "You can directly speak out your question~", as shown in S2001 in Figure 20 below.

[0276] After that, the target object can verbally describe his question. As shown in S2101 in Figure 21 below, the target object asks "Why is 'can' used here?". After the target object releases the finger, the selected target content and the second question described by the target object will be sent to the AI together, and the AI will feedback the second answer information and present Figure 21 the answer interface shown on the right. The answer interface includes the target content, as shown in S2102 in Figure 21 below, and also includes the second question input for the target content, as shown in S2103 in Figure 21 below, and the second answer information corresponding to the second question, as shown in S2104 in Figure 21 below.

[0277] In addition, it should be noted that the search information in the search information input area in the answer interface can also be updated to the second question, or updated to the target content + the second question, etc., which is not specifically limited here.

[0278] In the above embodiment, the object can perform multimodal AI search at any position, which can shorten the operation path of the object and realize the experience of being able to search anywhere.

[0279] Optionally, the answer interface can also include an intelligent Q&A entry, and the target object can switch modes based on this intelligent Q&A entry. One optional implementation method is as follows:

[0280] In response to a conversation operation triggered based on an intelligent Q&A entry, the client switches from the network search mode to the conversation search mode; then, the target object can ask questions about the second answer information, and the client responds to the question operation for the second answer information in the conversation search mode, presenting the third question input for the second answer information and the third answer information corresponding to the third question.

[0281] Still taking Figure 21 as an example, the intelligent Q&A entry in the answer interface is the "Continue to Ask AI" control shown in S2105. If the target object clicks S2105 to continue the conversation in the hotel list screened by AI, the network search mode can be switched to the conversational search mode. The target object can directly continue to ask questions based on the above second answer information. Then, AI combines the target object's personalized corpus (including favorited content, browsed content, etc.), the content retrieved from the network, and the target language model to give the answer (i.e., the third answer information) corresponding to the question (i.e., the third question). The specific implementation of AI Q&A is similar Figure 14 、 Figure 15 、 Figure 16 to the relevant parts, which will not be repeated here.

[0282] In the above implementation, on the result page returned by AI, the object can switch to the conversational Q&A by clicking "Continue to Ask AI". After switching to the AI conversation, each answer given by AI will be combined with the results of network search, the object's personalized corpus, and the target language model to answer the question.

[0283] It should be noted that the above Figures 18 to 21 listed related operations are optional and determined by the target object according to their own needs. The object can perform the above operations or not between two search operations. The above is just a simple example for illustration.

[0284] Suppose in a subsequent search, the target object queries another word, such as "naturally". When AI analyzes the target object's previous search content as context and determines that the two words ("artificial" and "naturally") queried belong to the same word learning and are relevant, that is, they meet the preset presentation conditions, then in this search process, AI sug, that is, the first search suggestion information, can be dynamically presented on the information search interface. By means of sug, it is prompted to the target object that the two search operations are related and that "artificial" and "naturally" are antonyms. After the target object becomes interested, they can click to view more detailed results of this search.

[0285] Such as Figure 22As shown, it is a schematic diagram of another type of first search suggestion information in an embodiment of the present application. Figure 22 That is, it means directly presenting the first search suggestion information "artificial and naturally are antonyms. It..." in the information search interface. The target object can click on this first search suggestion information to further view the corresponding search results, such as Figure 22 shown on the right side of the interface in the figure. This search interface summarizes some differences between artificial and naturally.

[0286] Such as Figure 23 As shown, it is a schematic diagram of yet another type of first search suggestion information in an embodiment of the present application. During this search process, AI sug can be presented in the Figure 23 shown information search interface to prompt the target object that artificial and naturally are antonyms. As shown in Figure 23 S2301 and S2302 in the figure, where the AI sug shown in S2301 represents the current processing progress, that is, it is in the stage of ongoing analysis; after the analysis is completed, the AI sug shown in S2302 can be presented. This AI sug is the first search suggestion information in an embodiment of the present application, which prompts the target object with the summary information of the current search results.

[0287] In an embodiment of the present application, the target object can click on the AI sug shown in S2302 to enter the search result interface and present AI search results, that is, some differences between artificial and naturally.

[0288] Compared with Figure 22 in Figure 23 the shown information search interface, there is an additional search suggestion information of "analyzing..." which more vividly reflects that the AI sug is dynamically presented along with the analysis process.

[0289] In the above implementation, when the object does not discover the association between search terms, the AI screening can actively provide the object with the associated content between different search terms to help the object better master relevant knowledge.

[0290] In addition, the dynamic AI sug provides the object with a quick preview of the AI search results combined with the context. By browsing the AI sug, the object can determine whether to view the content generated by the AI combined with the context; and, the AI sug appears during the object's search process and can prompt the object with the answer given by the AI at the first time when the object inputs; in addition, the form of the AI sug does not affect the object's browsing of the search term completion suggestions given by the web search, which can facilitate the object to switch between web search and AI search.

[0291] It should be emphasized here that in the above-mentioned embodiments, data such as the network search results and personalized corpus of the target object need to be obtained. The acquisition of these data has obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0292] Optionally, the information search method in the embodiments of the present application also supports intelligent summarization of the previously searched content by theme. When the object searches for content on multiple different themes, the target language model can be used to summarize by theme, and multiple AI sugs and their corresponding AI summary contents are given. An optional implementation method is as follows:

[0293] In response to the summary operation triggered by the target object, the client presents at least one second search suggestion information in the information search interface. Each second search suggestion information corresponds to a search theme, and the second search suggestion information is the summary information corresponding to the search theme; the search theme is obtained by dividing multiple search information related to the target object.

[0294] On this basis, the target object can view the summary content corresponding to any second search suggestion information. Specifically, in response to the selection operation of the target search suggestion information in at least one second search suggestion information, the client presents a search result summary interface corresponding to the target search suggestion information. The search result summary interface includes the summary content of the search results corresponding to each search information under the corresponding search theme.

[0295] Such as Figure 24 shown, which is a schematic diagram of a content summary process in the embodiments of the present application. When the target object inputs "summary" in the search information input area (such as Figure 24 the search box) of the information search interface, the AI will classify and summarize the content queried by the target object within the memory range and present it in the form of multiple sugs. As Figure 24 shown in S2401 and S2402 in

[0296] the target object has queried the travel guide of City A and multiple words. When the target object inputs "summary", the AI will summarize the content of the two themes respectively, and the two second search suggestion information presented are S2401 "Travel Guide for City A with 4000 yuan" and S2402 "Words Memorized Today".

[0297] Such as Figure 24As shown in interface S2403, it represents an example after the target object clicks on S2401, presenting a travel guide for City A. Moreover, the search information in the search box on this interface S2403 is updated to "Summary of the itinerary in City A" + "32 related collections".

[0298] Optionally, when the search topic is a travel guide, the corresponding summary content is a travel guide that includes a travel itinerary corresponding to at least one travel time period, such as the itinerary plans for the first day, the second day, and the third day listed in S2403. For example, on the first day, visit scenic spot x4 -> B Palace -> scenic spot c, and plan according to time periods such as morning, afternoon, and evening, listing specific itinerary plans. Similarly, specific itinerary plans are also given for the second day and the third day. For details, please refer to Figure 16 , which will not be elaborated here one by one.

[0299] In addition, the summary content is also marked with the collection content referred to in the travel guide, such as "Content reference from marked collections: x1 Historical Block, B Palace Museum, P1 Resort Hotel" marked at the bottom of S2403, etc.

[0300] Such as Figure 25 shown, it is a schematic diagram of another content summary process in the embodiment of the present application. When the target object inputs "summary" in the search information input area (such as the search box of Figure 25 ) on the information search interface, the AI will classify and summarize the content queried by the target object within the memory range and present it in the form of multiple sugs, as shown in S2401 and S2402 in Figure 25 .

[0301] If the target object selects to click on S2402 to view the words learned today, the AI will summarize all the words queried by the object today and present the summary content shown in S2404.

[0302] Optionally, if the search topic is knowledge learning, the summary content includes each knowledge point divided according to the importance level, and a knowledge example integrating each knowledge point.

[0303] For example Figure 25 the new words, phrases, and examples learned today listed in S2404 in, with the importance level ranging from one star to three stars. In addition, all the queried words can be combined into a paragraph of English, so that the target object can master all the words learned today by only memorizing one paragraph of English. As shown in the S2405 part of Figure 25 , the 6 words queried today are intelligently integrated to generate a knowledge example shown in S2405.

[0304] In the above embodiments, through intelligent summarization by sub-topics, it is possible to help the object classify and summarize different types of content. Moreover, for learning objects, this method can be used to organize the knowledge searched at different stages; for objects dealing with complex problems in guides, when the entire guide task is interrupted halfway, the functions achieved by this method can be used to summarize at any time, and by continuing the function of conversing with the AI, the formulation of the guide can be continued.

[0305] It should be emphasized here that in the summary scenario of the above-mentioned embodiments, it is necessary to summarize the content searched by the target object within a certain time range. The acquisition of these data involved in this process has obtained the permission or consent of the object, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0306] In addition, it should be noted that the above-mentioned travel scenario and word query scenario are only simple examples. In addition, other search scenarios are also applicable to the embodiments of the present application. For example, the target object can search for historical figures, hot news, film and television works, etc., which will not be elaborated here one by one.

[0307] It should be noted that the above is an introduction to the information search method in the embodiments of the present application from the client side. The following further explains the information search method in the embodiments of the present application from the server side:

[0308] Refer to Figure 26 As shown, it is a flowchart of the implementation of another information search method provided by the embodiments of the present application. Taking the server as the execution entity as an example, the specific implementation process of this method is as follows S261~S263:

[0309] S261: The server obtains the context corpus information of the target object. The context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0310] In the embodiments of the present application, data can be continuously collected according to each search process of the target object to update the context corpus information of the target object.

[0311] In the embodiments of the present application, the context corpus information of the target object can be in the form of a database, that is, a personalized corpus. The construction of this personalized corpus is roughly divided into three processes: A. Data collection, B. Data preprocessing, C. Construction of the corpus.

[0312] The following briefly explains the above three processes respectively:

[0313] A. Data collection.

[0314] The context corpus information is specifically divided into three categories: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0315] Among them, the content browsed by the target object in each search result and the content of the feedback behavior implemented by the target object can both be used as the favorite content of the target object.

[0316] In the embodiment of the present application, the collection method of search information is relatively simple and will not be elaborated here. The following briefly describes the collection process of the favorite content of the target object:

[0317] Specifically, when collecting the content browsed by the target object in each search result, each content browsed by the target object can be collected, or screening can be performed with reference to the browsing duration. For example: if the target object browses a certain web page content for more than the limited duration, taking 5s as an example, then collect this web page content.

[0318] In addition, taking the feedback behavior as collection as an example, when the target object collects an article, collect this article; when the target object collects a partial content of an article (such as words, phrases, examples, paragraphs, etc.), collect this partial content. Of course, the collection methods of other feedback behaviors except collection are the same as the collection behavior, and the repeated parts will not be elaborated.

[0319] The following briefly describes the specific process of collecting the partial content when the target object collects a partial content of an article:

[0320] In the embodiment of the present application, the method for the target object to collect partial content is not specifically limited. For example, it can be automatically selecting the partial content after double-clicking. Among them, the technical solution for selecting the partial content is implemented based on the text selection engine of the client (subsequently, the client is taken as the browser for example).

[0321] Taking the partial content as a certain word or a paragraph of text as an example, when the target object double-clicks a certain word or a paragraph of text, the browser will automatically select the entire word or the entire paragraph of text according to the selection range of the target object. And, the target object can change the selected content by adjusting the selection edge up and down. The specific implementation process is as follows, including the following steps Sa1~Sa7:

[0322] Sa1: When the target object double-clicks a certain word or a paragraph of text, the browser calculates the start position and end position of the selected text according to the selection range of the target object.

[0323] Sa2: The browser determines the range of the selected text according to the start position and end position.

[0324] Sa3: The browser passes the range of the selected text to the text selection engine.

[0325] Sa4: The text selection engine traverses all text nodes in the document tree according to the range of the selected text and finds the text node that contains the selected text.

[0326] Sa5: The text selection engine locates the range of the selected text in the text node and marks the style of the selected text.

[0327] Sa6: The browser applies the style of the selected text, such as the style of selecting an entire paragraph, to the document so that the target object can see the selected text.

[0328] Sa7: If the target object touches the edge of the box for selecting an entire paragraph and moves, judge the end point of the movement of the target object. According to the distance that the target object moves, use it as the new selection range of the target object, and recalculate the start position and end position of the text.

[0329] It should be noted that the above-listed is only an optional implementation manner for selecting partial content of a document according to the selection range of a target object in the embodiments of the present application. Other related manners are also applicable to the embodiments of the present application and will not be elaborated one by one here.

[0330] B. Data preprocessing.

[0331] In the embodiments of the present application, when constructing the context corpus information of the target object by collecting the above data, the above-collected data can also be preprocessed. The specific implementation manner is as follows, including the following steps Sb1 to Sb4:

[0332] Sb1. Remove HTML tags: Use methods such as an HTML parser or regular expressions to remove HTML tags in the web page content and only retain the text content.

[0333] Sb2. Sentence splitting: Split the text content into sentences.

[0334] Optionally, it can be implemented using a sentence tokenizer. Common methods are rule-based or machine learning models. Here is just a simple example and no specific limitation is made in this article.

[0335] Sb3. Word segmentation: Split the sentence into words or phrases.

[0336] Optionally, a word segmentation tool such as the Natural Language Toolkit (NLTK), spaCy, etc. can be used for word segmentation processing. Here is just a simple example and no specific limitation is made in this article.

[0337] Sb4, Part-of-speech tagging: Tag each word with its part of speech, such as noun, verb, adjective, etc.

[0338] Optionally, a part-of-speech tagger can be used to achieve this. Common methods are rule-based or machine learning models. Here is just a simple example, and this article does not make specific limitations.

[0339] C. Construct a corpus.

[0340] After preprocessing the collected data through the above methods, the context corpus information corresponding to the target object, also known as the personalized corpus, can be constructed. The specific implementation method is as follows:

[0341] Construct the preprocessed data into individual text samples and store them in a corpus. Each text sample can be the content of a location, a word, a phrase, an example sentence, a paragraph, an article, or a web page, etc.

[0342] Specifically, these text samples can be stored in a suitable text format, such as txt, csv, etc. This article does not make specific limitations.

[0343] S262: After the server receives a search request triggered by the target search information, if it determines that the preset presentation condition is met, it inputs the target search information and the context corpus information into the target language model to generate the current search result and the summary information of the current search result; the preset presentation condition is that the current search operation of the target object is related to the historical search operations of the target object.

[0344] Refer to Figure 27 As shown, it is a schematic flowchart of an AI multi-round search implementation in network search in an embodiment of this application.

[0345] Specifically, after the target object enters a query in the search engine, it can obtain the search results returned by the search engine; then, the target object browses and collects the search results and stores these contents as the personalized corpus of the object; after that, the above query and the personalized corpus can be used as the context, combined with the object's new question, and input to the target language model together. The target language model combines the above contents and gives a new answer.

[0346] In the embodiments of the present application, by combining with a target language model, the search terms in the target object's previous web search, the content of the pages browsed after the search, the content corresponding to the feedback behaviors such as the target object's active collection and like of the pages, etc. are used as context corpus information to create a personalized corpus, and the context corpus information generated each time a search is performed is provided to the target language model. When the target object asks a question again, the target language model will, based on the obtained content, give a summary of the AI answer in the sug page (i.e., the information search interface). When the target object clicks on the sug of the AI, the search answer generated by the target language model by combining the personalized corpus information of the target object can be seen, providing the target object with an experience where a conventional web search becomes a multi-round search.

[0347] Among them, the above summary information is the first search suggestion information, which is a dynamic AI sug. When the target language model determines that it is related to the previous search behavior of the target object, during the process of the target object's search input, the sug will appear dynamically. The other entries in the sug page are still the sug completions for the current query, and the target object can still perform a web search. That is, the target object can freely choose web search and target language model search. The specific generation process of this information is as follows Sc1~Sc5:

[0348] Sc1. Context understanding: After the target object inputs a question (referring to the target search information), the target language model combines the question of the target object and the obtained personalized corpus (also called context), understands the intention of the target object according to the context, and generates a suitable answer (i.e., the result of this search).

[0349] Specifically, this process usually needs to encode the question into a form that the model can understand, such as word embedding or vector representation.

[0350] Sc2. Answer generation: The target language model generates an answer according to the context.

[0351] Specifically, it can be a generation algorithm based on the model, such as a recurrent neural network (RNN) or a transformer, etc., to generate a suitable sug according to the semantic and syntactic rules of the context.

[0352] Sc3. Generation of suggestions for continued questions: While generating the answer, the target language model will also generate some possible answer sugs, and this process is usually based on the generation algorithm and the output of the model.

[0353] Specifically, this generation algorithm can select the most suitable several from the candidate statements as sugs according to the probability distribution generated by the model or other rules.

[0354] Sc4. Suggestion Sorting and Filtering: The generated suggs need to be sorted and filtered to provide the most relevant and useful suggestions for the target object.

[0355] Specifically, this sorting can be achieved based on some metrics, such as the confidence of the probability distribution, semantic relevance, etc. Filtering can consider some rules or heuristic methods, such as grammar rules, common question patterns, etc.

[0356] Sc5. According to the change of the query input by the target object, the suggs generated by the AI each time also change.

[0357] As described above Figure 7 in S702 or Figure 22 etc. are taken as examples, which are examples of several first search suggestion information listed in the embodiments of the present application. For details, refer to the above embodiments, and repeated parts will not be elaborated.

[0358] In the embodiments of the present application, the search results generated this time by the target language model in combination with the target search information and the context corpus information are AI personalized filtering results that meet the needs of the object. Specifically, this result is generated in the following way:

[0359] For the target object, the target object can conduct multiple network searches. The target language model can cross-filter the previous search terms of the target object to screen the search results for the target object this time, and display them in a suitable form according to the content types of the screened content; or compare multiple search terms with each other and give the results after comparison.

[0360] Still taking the above-listed travel scenario and word query scenario as examples below, the specific generation processes of the corresponding search results this time will be elaborated respectively.

[0361] (1) Travel scenario.

[0362] In the travel scenario, the AI cross-filters by comparing the previous search terms of the target object and combining the locations collected before, etc., to provide search results that better meet the preferences of the target object. An optional implementation method is as follows:

[0363] If the target search information is related to the geographical location, the search results this time can be generated in the following way:

[0364] First, determine the collection locations related to geographical locations in the collection content of the target object; then, input the collection locations into the target language model, use the target language model to learn the feature information of the collection locations, and combine the historical behavior of the target object to determine at least one recommended location for the target object; finally, for each recommended location, collect the content related to the recommended location in the network, extract the basic information of the recommended location, and at least one associated web result, and use them as the search results for this time.

[0365] Still taking Figures 3 to 16 the listed travel scenario as an example, when the target object searches for hotels in City A for the second time, the AI uses the locations collected by the target object in the personalized corpus as the screening basis to find hotels near the location for the target object. The specific process is as follows Sd1~Sd4:

[0366] Sd1. Input the location information collected by the target object into the target language model to let the target language model learn the features of these locations.

[0367] Sd2. Recommend hotels: Use a collaborative filtering-based or content-based recommendation algorithm to recommend nearby hotels.

[0368] Specifically, the collaborative filtering-based recommendation algorithm means: recommending items based on the historical behavior of the object and the behavior of other objects.

[0369] For example, if both Object A and Object B like the hotels near a certain location, then the system will recommend this hotel to Object C.

[0370] Specifically, the content-based recommendation algorithm means: recommending items based on the attributes of the items and the historical behavior of the target object.

[0371] For example, if the target object likes the hotels near a certain location, then the system will recommend other similar hotels to the target object.

[0372] Sd3. Screen articles on the network: After recommending hotels, use web crawler technology to collect articles related to the recommended hotels on the network, such as reviews, ratings, pictures, etc. Then, use natural language processing technology to analyze and process these articles, extract the useful information from them as tips for the target object. And obtain the network original text connection so that the target object can view the original text.

[0373] Sd4. Combine map display: Use the map application programming interface (Application Programming Interface, API) to display the collected locations and the recommended hotels on the map.

[0374] Still taking Figure 8Taking the example shown below, where S801 is the map module presented using the map API, including the collected locations and recommended hotels; in addition, as shown in Figure 9 the parts S901, S902, and S903 shown are articles related to the recommended hotels collected from the Internet. For specific details, please refer to the above embodiments, and repeated parts will not be elaborated here.

[0375] (2) Word query scenario.

[0376] In the word query scenario, the AI lists the similarities and differences between the previous search terms and the current search term for the target object by comparing the previous search terms of the target object. One optional implementation method is as follows:

[0377] If the target search information is related to knowledge point learning, the following method can be used to generate the search result for this time:

[0378] First, input the knowledge point to be searched corresponding to the target search information into the trained knowledge point recognition model; then, based on the knowledge point recognition model, if it is determined that the associated knowledge points of the knowledge point to be searched have been searched by the target object, then the knowledge point to be searched, the determined associated knowledge points of the knowledge point to be searched, and the relevant explanations are used as the search result for this time.

[0379] The following takes the knowledge point to be searched as a word and the knowledge point recognition model as an antonym recognition model as an example to briefly explain the above method. The specific process is as follows Se1~Se6:

[0380] Se1. Data collection: First, it is necessary to collect corpus data containing a large number of words and their antonyms. This data can come from existing antonym dictionaries, synonym dictionaries, corpora, or manually annotated data sets.

[0381] Se2. Model construction: Use the collected data to train an antonym recognition model.

[0382] Specifically, machine learning algorithms such as support vector machines (SVM) or deep learning models such as RNN or Transformer models can be used. This article does not make specific limitations.

[0383] Se3. Feature extraction: For each word, it is necessary to convert it into a feature representation that can be processed by machine learning algorithms.

[0384] Specifically, a word vector model (such as Word2Vec or GloVe) can be used to convert the word into a vector representation, or other feature engineering methods can be used. This article does not make specific limitations.

[0385] Se4. Training Model: Use the collected data and extracted features to train an antonym recognition model. The supervised learning method can be used, with words and their antonym labels as training samples.

[0386] Se5. Model Evaluation and Optimization: Evaluate the trained antonym recognition model using the test dataset, and optimize and adjust the antonym recognition model according to the evaluation results.

[0387] Se6. Applying the Model: Apply the trained antonym recognition model to actual scenarios.

[0388] Specifically, when the target object queries a word, the antonym recognition model can determine whether its antonym has been input by the target object. If it has been input, then this group of words and their related explanations are returned together.

[0389] Still taking Figure 22 shown as an example, where the artificial queried by the target object and the previously queried naturally are antonyms, and the corresponding related explanations are as shown on the right interface in Figure 22 For details, refer to the above embodiments, and repeated parts will not be elaborated.

[0390] Optionally, the search result interface further includes additional prompt information matching the current search results. In the embodiments of the present application, the additional prompt information can be determined and fed back to the client in the following manner, and then presented to the target object by the client. The specific process is as follows:

[0391] First, it is necessary to extract the date information and location information during the current search process; then, based on the date information and location information, generate the prompt information corresponding to each preset prompt template; then, sort the prompt information corresponding to each prompt template according to their respective preset weights; finally, use the prompt information within the specified order range of the sorting result as the additional prompt information.

[0392] Specifically, the server side can store two types of prompt templates: a date information-containing template and a date information-free template.

[0393] Among them, the date information-containing templates include but are not limited to: a weather template, b major festival template, c travel restriction template, d travel clothing template; the date-free templates include but are not limited to: e venue restriction information template (such as scenic spot visit restriction information template), f merchant welfare information template.

[0394] Among them, the content included in each prompt template is as follows:

[0395] The a weather template includes: the temperature and weather conditions at the local area on the date.

[0396] The major festival template includes: the date and local festival activity information within the date range;

[0397] The travel restriction template includes: the date and local traffic restriction information within the date range;

[0398] The travel clothing template includes: the travel clothing suggestion information combined with the temperature within the date range;

[0399] The restriction information template of the venue includes: the restricted entry information of the relevant location;

[0400] The welfare information template of the merchant includes: the welfare information of the relevant location.

[0401] In the embodiment of the present application, when the target object views the AI intelligent screening result, the date information and location information in the relevant search process of the target object can be extracted, and the date information and location information are input into the search engine to obtain the template information of a-f in sequence. Furthermore, the obtained template information is sorted according to the weights of a-f. Taking the top 2 in the specified order range as an example, the top 2 templates in the sorting can be displayed on the final search result interface; when there are less than 2, 1 is displayed; when there is less than 1, it is not displayed.

[0402] Among them, the preset weights corresponding to the prompt templates can be flexibly set according to actual needs. For example, the weights of a-f can be set to decrease in sequence, or the weights of a-f can be set to increase in sequence, etc., which are not specifically limited in this article.

[0403] Specifically, the additional prompt information such as Figure 10 S1001 and S1002 in Figure 11 S1101 in Figure 12 S1201 in Figure 13 S1301 shown in

[0404] In the above implementation manner, the presentation of the additional prompt information helps the object to understand more content at one time, reduces the search path of the object, and further improves the information search efficiency of the object this time.

[0405] S263: The server feeds back the summary information and the current search result to the client, so that after the client determines that the preset presentation condition is met, the first search suggestion information including the summary information is presented in the information search interface, and after responding to the viewing operation triggered by the first search suggestion information, the search result interface including the current search result is presented.

[0406] Such as Figure 7 、 Figure 8 、 Figure 22 、 Figure 23As shown, it is a schematic diagram of several first search suggestion information and corresponding search result interfaces listed in the embodiments of the present application. For the specific implementation manner, reference can be made to the above embodiments, and repeated parts will not be elaborated.

[0407] In addition, the information search method in the embodiments of the present application also supports switching the AI screening result to the conversational search mode. For example, it is set that the search result interface or the answer interface further includes an intelligent Q&A entry, and the target object can switch the mode based on this intelligent Q&A entry.

[0408] In the embodiments of the present application, the answer information generated in the AI conversation is essentially the search result combined with the personalized corpus. After the target object switches to the AI conversation, each time the AI answers, it will answer the question in combination with the results of network search, the target object's personalized corpus, and the target language model. The specific process is as follows Se1~Se3:

[0409] Se1: Establish an index for the target object's personalized corpus.

[0410] In the embodiments of the present application, after the personalized corpus has been established and data cleaning and preprocessing have been performed on it, an index needs to be established. This index is a data structure used to quickly retrieve text data.

[0411] Optionally, an open-source search engine library such as Elasticsearch, Solr, etc. can be used to establish an index for the text data in the personalized corpus.

[0412] Se2: Develop a search plugin. Develop a search plugin for retrieving the text data in the personalized corpus.

[0413] Specifically, this plugin can use the API of the search engine library to retrieve the text data in the personalized corpus through keywords and integrate the search results into the answer.

[0414] Se3: Integrate it into the target language model: Integrate this search plugin into the target language model so that the plugin can be called when answering questions and the content of the personalized corpus can be displayed.

[0415] In the embodiments of the present application, in the AI conversation process, combining the personalized corpus with the technical solution of the plugin can make the content generation of the target language model more flexible. In addition, the present application also supports that the object can choose to turn on or off the plugin to affect the generated content.

[0416] Optionally, this application also supports multimodal AI search for any content by the user. In any scenario on the browsing result page, the user can long-press to select text or an image, and then ask questions about the selected content. The target language model will combine the selected content and the user's question to provide an answer. An optional implementation is as follows:

[0417] After the target user selects content from the search results, determine the selected target content based on the selection range of the target user. Then, detect a voice input event. After detecting the voice input event, convert the voice of the question input by the target user for the target content into text. Then, perform semantic analysis and keyword extraction on the text to obtain a text recognition result. Finally, input the text recognition result and the target content into the target language model to obtain the answer information generated by the target language model for the target content, and feedback it to the client, so that the client presents a corresponding answer interface in response to the question operation for the target content. The answer interface includes the target content, the question, and the answer information.

[0418] Taking the client as a browser as an example, the process specifically includes the following steps Sf1 to Sf6:

[0419] Sf1. When the target user long-presses on a word or a block of text, the browser calculates the start and end positions of the selected text based on the selection range of the target user, and then locates and selects this block of text.

[0420] Sf2. Detect a voice input event: Detect a voice input event in the browser and be able to respond in a timely manner when the target user starts voice input.

[0421] Sf3. Call the voice recognition API: When the target user starts voice input, call the voice recognition API provided by the mobile operating system to convert the voice input of the target user into text.

[0422] Sf4. Process the recognition result: The voice recognition API will return the recognized text, and process the text, such as performing semantic analysis and keyword extraction, to more accurately match the search results.

[0423] Sf5. Provide the text obtained in step Sf1 and the question obtained in step Sf4 to the target language model together.

[0424] Sf6. Use the target language model to give an answer based on the input content.

[0425] As Figure 20 、 Figure 21 shown, for details, please refer to the above embodiments, and the repeated parts will not be elaborated again.

[0426] In the above embodiments, the multi-modal AI search that supports objects at any position can shorten the path of object operations and achieve the experience of being searchable anywhere.

[0427] It should be noted that in the embodiments of the present application, during any search process of the target object, the target object may also be interested in the local content of the corresponding search results. For such content, AI intelligent collection can be performed.

[0428] When the target object collects a certain paragraph of text, the location, name, and other content in this part of the content will be highlighted and specially displayed. For example, the above Figure 6 、 Figure 18 、 Figure 19 etc. are several related examples of AI intelligent collection listed in the embodiments of the present application. For the specific implementation method, reference can be made to the above embodiments, and repeated parts will not be elaborated.

[0429] The following briefly describes the specific implementation scheme of AI intelligent collection:

[0430] When the content collected by the target object is a certain paragraph, the content in this paragraph can be tagged from the following several dimensions:

[0431] (1) Location tag, (2) Event tag, (3) Item tag, (4) Time tag.

[0432] The following separately describes the tagging methods for the above four dimensions:

[0433] (1) The tagging method for the location tag is as follows:

[0434] In the embodiments of the present application, locations can be divided into the following categories: A restaurant, B scenic spot, C store location, D hotel location, E other locations. Specifically, tags need to be added to these categories of locations in the content, and the number of occurrences of repeated locations needs to be marked.

[0435] Specifically, there are many methods for tagging the above categories of locations and marking the number of occurrences. The following briefly lists two of them:

[0436] Method 1: Train a new location model. The specific implementation scheme of this method includes the following steps Sg1 to Sg6:

[0437] Sg1. Data preparation: Collect the strategy data of each travel location and organize it into a text format.

[0438] Optionally, a crawler tool can be used to crawl the strategy data from travel websites or obtain it from other data sources.

[0439] Sg2. Data cleaning: Clean the collected strategy data to remove useless information and noise data.

[0440] Optionally, natural language processing techniques can be used to process the strategy text through word segmentation, part-of-speech tagging, named entity recognition, etc.

[0441] In the embodiments of the present application, preprocessing such as word segmentation and part-of-speech tagging of the strategy text is carried out to facilitate better understanding of the structure and semantics of the strategy text.

[0442] Among them, named entity recognition (NER) can be specifically implemented using a named entity recognition model, such as a conditional random field (CRF) based on machine learning or a recurrent neural network (RNN) model of deep learning, to perform entity recognition on the article. These models will label named entities such as locations, people's names, and organizations in the strategy text.

[0443] Sg3. Location type classification: Use machine learning or deep learning algorithms, such as Naive Bayes, support vector machines, convolutional neural networks, etc., to classify the locations in the strategy text. The locations are divided into A restaurants, B scenic spots, C store locations, D hotel locations, and E other locations, and a classification model is trained for each type.

[0444] Sg4. Model evaluation and tuning: Use the test set to evaluate the trained classification model, and the evaluation metrics can include accuracy, recall rate, F1 value, etc. Furthermore, according to the evaluation results, the classification model is tuned, such as adjusting model parameters, increasing the amount of training data, etc.

[0445] Sg5. Location marking of the content collected by the target object: Input the strategy text into the tuned classification model, and the classification model will output the label of the location type and the occurrence times information of the repeated scenic spots.

[0446] Sg6. Display and storage of the collection result: Visualize the output result.

[0447] Specifically, compare the recognized location nouns with the original text to determine the position where they appear in the original text, and display a special style at that position, such as Figure 5 shown in S502 of

[0448] In addition, in the animation display of the collection process of the target object, different types of locations can be marked with different colors; also, in the collection folder of the target object, the corresponding label can be displayed for each type of location, and the occurrence times can be marked next to the repeated scenic spots.

[0449] Method 2: Information comparison. The specific implementation scheme of this method includes the following steps Sh1 to Sh4:

[0450] Sh1. Location Word Extraction: For the text collected by the target object, use natural language processing techniques, including word segmentation, part-of-speech tagging, named entity recognition, etc., to process the strategy text and extract the location words in it.

[0451] Specifically, open-source tools such as Jieba word segmentation and Stanford Named Entity Recognition (StanfordNER) can be used.

[0452] Sh2. Location Classification Recognition: For each location word, use network retrieval or comparison in existing data to determine which location classification it belongs to among A restaurants, B scenic spots, C store locations, D hotel locations, and E other locations.

[0453] Specifically, the search engine can be used to search for the location name and determine its type based on the information in the search results. Existing location classification data, such as map software, can also be used for comparison.

[0454] Sh3. Duplicate Scenic Spot Recognition: For locations of the same type, use techniques such as similarity calculation to determine duplicate scenic spots in the strategy text.

[0455] Specifically, the location words can be calculated for similarity, such as cosine similarity and Jaccard similarity, to determine whether they are similar. If they are similar, they are considered duplicate scenic spots and the number of occurrences is recorded.

[0456] Sh4. Collection Result Display and Storage: Visualize the output results.

[0457] Similarly, compare the identified location nouns with the original text to determine their positions in the original text and display a special style at that position, such as Figure 5 shown in S502 of

[0458] In addition, in the animation display of the target object's collection process, different types of locations can be marked with different colors; in the target object's collection folder, corresponding labels can be displayed for each type of location, and the number of occurrences can be marked next to the duplicate scenic spots.

[0459] In summary, the method of retraining the model in Method 1 above can obtain higher accuracy and customization capabilities. The above Scheme 2 can quickly implement the function through the method of information comparison.

[0460] (2) The method of tagging event labels is as follows:

[0461] In the embodiments of the present application, the definition of an event includes two categories: events on specific dates (such as public events like historical anniversaries, festival celebrations, etc.), and non-specific date events (such as daily float parades, etc.).

[0462] Specifically, the method of event tagging is similar to that of location tagging. A new event model can be trained to use the model to tag major events that appear in the content collected by the target object; or the method of information comparison can be used to compare network information to identify events in the content.

[0463] (III) The method of item tagging is as follows:

[0464] In the embodiments of the present application, the definition of an item includes but is not limited to special products, special foods, special toys, etc.

[0465] Specifically, the method of item tagging is also similar to that of location tagging. A new item model can be trained to use the model to tag items that appear in the content collected by the target object; or the method of information comparison can be used to compare network information to identify items in the content.

[0466] (IV) The method of time tagging is as follows:

[0467] In the embodiments of the present application, a large language time information recognition model can be used to recognize the time information in the text, including but not limited to time points, dates, seasons, years, etc. Furthermore, using this model, it is determined whether there are associated scenic spots, events, and items at the time point in the text; for the associated scenic spots, events, and items, corresponding time tags are added.

[0468] It should be noted that the above is an example where the content collected by the target object is a certain paragraph. When the content collected by the target object is a word (or phrase, example sentence, etc.), the following method can be used to tag the word (or phrase, example sentence, etc.):

[0469] If the target object selects a word (or phrase), the word (or phrase) can be compared with the existing database to obtain the difficulty score of the word (or phrase); furthermore, the score of the word (or phrase) is displayed in the relevant interface (such as the search result interface, collection interface, etc.) in the form of a tag.

[0470] If the target object collects a sentence, the words in the sentence can be scored one by one, and the average score is calculated as the score of the sentence; furthermore, the words, the sentence, and the corresponding difficulty scores are stored together, and the score of the sentence is displayed in the relevant interface (such as the search result interface, collection interface, etc.) in the form of a tag.

[0471] It should be noted that the above is only a simple example with the content collected by the target object being paragraphs, words, phrases, sentences, etc. In addition, the content can also be in other forms, and the specific analysis method is similar, so the repeated parts will not be elaborated.

[0472] Optionally, the information search method in the embodiments of the present application also supports intelligent summarization by theme for the previously searched content. When the object searches for content on multiple different themes, the target language model can summarize by theme and give multiple AI sugs and their corresponding AI results. An optional implementation method is as follows:

[0473] The client can send a summary request to the server in response to the summary operation of the target object. After receiving the summary request triggered by the target object, the server respectively determines the search themes corresponding to multiple search information related to the target object; furthermore, through the target language model, taking the search theme as a unit, the search results of different search themes are summarized to obtain the summary content corresponding to each search theme; furthermore, the obtained summary content is scored and sorted, and the summary content with a score lower than the preset score threshold is filtered out; finally, the server generates the second search suggestion information corresponding to each remaining summary content through the target language model and feeds back the generated second search suggestion information to the client, so that the client presents at least one second search suggestion information in the information search interface in response to the summary operation, and presents the search result summary interface corresponding to the target search suggestion information after responding to the selection operation of the target search suggestion information in at least one second search suggestion information. The search result summary interface includes the summary content under the corresponding search theme; each second search suggestion information is the abstract information corresponding to the corresponding search theme.

[0474] The specific technical solutions are as follows Sn1~Sn5:

[0475] Sn1. Determine different themes.

[0476] Optionally, the following method can be used to determine the different theme contents in the query input by the target object:

[0477] A. Keyword matching: The target language model can extract the keywords in the current question (i.e., the target search information) and match them with the keywords in the context (i.e., the context corpus information). If there are keywords related to the current question in the context corpus information, then this part of the context may be useful.

[0478] B. Semantic similarity: The target language model can calculate the semantic similarity between the current question and the context. By comparing the semantic representations of the question and the context, it can be judged which context has a stronger semantic connection with the current question, so as to determine its usefulness.

[0479] C. Topic Modeling: The target language model can use topic modeling techniques, such as Latent Dirichlet Allocation (LDA) or pre-trained language representation models like Bidirectional Encoder Representations from Transformers (BERT), etc., to identify the topics in the context. If the topics in the context are highly relevant to the topic of the current question, then this part of the context may be useful.

[0480] Sn2. Tag the content by topic, and the content belonging to the same search topic is tagged with a unified tag;

[0481] Sn3. When the target object inputs "summary", the target language model summarizes according to the different topic contents of the target object. Taking the search topic as the unit, summarize the different topic contents.

[0482] Sn4. Score and rank, score and rank the generated summary content, and filter out the content with lower scores;

[0483] The following are some common scoring factors:

[0484] A. Keyword Matching: The search engine checks whether the keywords on the web page match the search query of the target object. If the keywords appear frequently in the title, body, and other tags, then the web page may get a higher score.

[0485] B. Content Quality: The search engine evaluates the content quality of the web page, including the accuracy, integrity, and relevance of the information. Web pages with higher content quality usually get higher scores.

[0486] C. External Links: The search engine considers the number and quality of links to the web page from other websites. If other trusted websites link to the web page, then the web page may get a higher score.

[0487] D. Target Object Experience: The search engine analyzes the interaction between the target object and the search results, such as click-through rate, dwell time, and bounce rate, etc. If the target object is interested in a search result and interacts with it, then the web page may get a higher score.

[0488] E. Web Page Structure and Markup: The search engine analyzes the structure and markup of the web page, such as title tags, paragraph tags, and image tags. Good web page structure and markup can improve the search engine's understanding of the web page content, thus affecting the score.

[0489] It should be noted that the above - listed several ways of scoring and ranking the generated summary content are only simple examples. In addition, other scoring methods are also applicable to the embodiments of this application and will not be elaborated one by one here.

[0490] Sn5. Generate a corresponding AI sug (i.e., the second search suggestion information in this article) for each summary content. The content of the AI sug is the abstract of the corresponding content, and the abstract contains the keywords of the theme.

[0491] Such as Figure 24 、 Figure 25 As shown, for details, refer to the above - mentioned embodiments, and repeated parts will not be elaborated.

[0492] In addition, it should be noted that in the embodiments of this application, the above - collected content can be used to generate summary content. Specifically, the collected content of the target object can be classified according to a preset dimension, and classification labels included in each piece of collected content are marked (for details, refer to the above - mentioned embodiments, such as marking location labels, event labels, item labels, time labels, etc.); furthermore, when using the target language model to summarize the search results of different search topics in units of search topics to obtain the summary content corresponding to each search topic, the target collected content related to the target search information in the collected content can be retrieved according to the classification label; then, through the target language model, in units of search topics, combined with the target collected content, the search results of different search topics are summarized to obtain the summary content corresponding to each search topic.

[0493] The following takes the above - mentioned travel scenario and word - query scenario as examples for simple illustration:

[0494] (1) Travel scenario.

[0495] In the embodiments of this application, if the target search information is related to travel guides, the corresponding summary content, such as travel guides, can be generated by calling location labels and using the target language model. An optional implementation method is as follows:

[0496] First, analyze the collected content of the target object, and extract the collection locations related to the travel destination in the collected content of the target object; then, according to the extracted collection locations, collect the network information related to the collection locations; combine the target language model and the network information to generate recommended locations; finally, after forming a location library with the collection locations and the recommended locations, generate the corresponding travel guides according to each location in the location library.

[0497] Specifically, the calling scheme of the location labels involved in the above - mentioned implementation method is as follows Si1~Si6:

[0498] Si1. Favorite Content Management: In the target object's favorites, label each location with five location tags: A restaurant, B scenic spot, C store location, D hotel location, and E others, and mark the number of occurrences of duplicate scenic spots.

[0499] Si2. Extraction of Target Object's Favorite Content: When the target object requests to create a strategy for a specific scenic spot, the target language model (such as chatGPT, full name Chat Generative Pre-trained Transformer) needs to analyze the target object's favorite content and extract the favorite locations related to the destination.

[0500] Specifically, machine learning algorithms or rule engines can be used to analyze and extract the preferences of the target object.

[0501] Si3. Data Collection: According to the extracted relevant locations, collect relevant information from the network. Among them:

[0502] For type A restaurants, collect information such as the location of the restaurant, evaluation score, per capita consumption, special signature dishes, pictures, etc.;

[0503] For type B scenic spots, collect information such as the location of the scenic spot, evaluation score, ticket price, opening hours, pictures, etc.;

[0504] For type C store information, collect information such as location, evaluation score, opening hours, per capita consumption, pictures, etc.;

[0505] For type D hotels, collect information such as location, room type, price corresponding to the room type, room status information, pictures, etc.;

[0506] For type E others, collect information such as location, evaluation score, pictures, etc.

[0507] Si4. Exploratory Recommendation: Combine the target language model and network information to generate more locations based on the locations that the target object is interested in.

[0508] Optionally, the types generated in this step include but are not limited to the following two:

[0509] Method 1: According to the location information of the favorite locations, collect other travel locations within 1 km nearby.

[0510] Method 2: According to the type of the favorite locations, use a recommendation algorithm to collect scenic spots of related types.

[0511] Specifically as follows:

[0512] A. Based on the Class A restaurant information in the target object's preferences, collect scenic spots, shops, and hotels within 1 km of the restaurant, and prioritize the locations that the target object has already favorited. Then, use a recommendation algorithm to find other restaurants similar to Class A restaurants.

[0513] B. Based on the address information of Class B scenic spots, collect other nearby locations, including restaurants, shops, hotels, and scenic spots, and prioritize the scenic spots that the target object has already favorited. Then, using technologies such as recommendation algorithms based on the Class B scenic spot information in the target object's preferences, generate recommendations for scenic spots that the target object may be interested in but has not yet learned about. For example, if the target object has favorited natural scenery scenic spots, the recommendation algorithm can recommend relevant nature reserves or wildlife parks, etc.

[0514] C. Based on the Class C shop information in the target object's preferences, collect scenic spots, restaurants, and hotels within 1 km of the shop, and prioritize the locations that the target object has already favorited. Then, use a recommendation algorithm to find other shops similar to Class C shops.

[0515] D. Based on the Class D hotel location in the target object's preferences, collect scenic spots, restaurants, and shops within 1 km of the hotel, and prioritize the locations that the target object has already favorited. Then, use a recommendation algorithm to find other hotels similar to Class D hotels.

[0516] E. Based on the Class E other locations in the target object's preferences, collect scenic spots, restaurants, shops, and hotels within 1 km of the location, and prioritize the locations that the target object has already favorited.

[0517] Si5. Establish a location library and sort: The target language model uses the five types of locations favorited by the target object, the locations generated by exploratory recommendations, and the strategy locations found on the Internet as the location library, and generates travel strategies based on the locations in the location library.

[0518] Among them, the priorities of the three sources are: locations favorited by the target object > locations generated by exploratory recommendations > locations recommended on the Internet. Among the three sources, the locations of the five location types are sorted according to their respective priorities. The higher the number of occurrences, the higher the priority.

[0519] Specifically, for the five types of locations listed above, the locations under each location type are sorted according to the corresponding source priorities. For example, for the three locations under Class A locations: Location A1, Location A2, and Location A3, which come from the locations favorited by the target object, locations generated by exploratory recommendations, and locations recommended on the Internet respectively, the sorting result is: Location A1 > Location A2 > Location A3. Specifically, for locations from the same source, they can be sorted randomly, sorted according to the number of occurrences, sorted according to the location name, etc., and no specific restrictions are made here.

[0520] Si6. Itinerary planning: Select 5 types of locations from the location library to generate a travel guide.

[0521] Optionally, the travel guide includes an itinerary plan corresponding to at least one travel time period; when generating the corresponding travel guide based on each location in the location library, the travel time set by the target object can be combined to divide at least one travel time period; furthermore, for each travel time period, according to the source priority corresponding to each location in the location library, select the corresponding number and corresponding categories of locations to generate an itinerary plan corresponding to the travel time period.

[0522] For example, in the generated guide, the itinerary for each day includes: 1 - 2 location of type B scenic spots, 3 location of type A restaurants, 0 - 2 location of type shops, 1 location of type D hotels, 0 - 2 location of type E others. Select the corresponding number and corresponding categories of locations according to the priority order in the location library, and combine with the specific number of days required by the target object, and combine algorithms such as route calculation of the target language model to generate a usage plan for the travel guide event tags.

[0523] Still taking Figure 16 、 Figure 24 S2403, etc. in

[0524] as an example, which are examples of several travel guides listed in the embodiments of the present application. For details, please refer to the above embodiments, and the repeated parts will not be elaborated.

[0525] Specifically, the calling scheme of the event tags involved in the above implementation manner is as follows Sj1 - Sj4:

[0526] Sj1. When the target object puts forward a guide requirement without specific time information, search for non-fixed date events in the favorites.

[0527] Sj2. Provide event suggestions for the target object in the form of tips for non-fixed date events related to the destination outside the accommodation itinerary.

[0528] Sj3. When the requirement of the target object includes specific time information, search for fixed date events in the favorites.

[0529] Sj4. Judge whether the travel date of the target object coincides with the fixed date event. When the dates coincide, arrange the event in the itinerary plan.

[0530] In an embodiment of the present application, it is also possible to search for item tags of relevant locations in the favorite content according to the travel guide requirements proposed by the target object; furthermore, based on the item tags, generate item suggestions for the target object outside the itinerary; and determine the shops where the items can be purchased through an online search, and add the target shop locations within a preset distance range from the formed route in the found shops to the formed corresponding itinerary plan.

[0531] Specifically, the usage scenarios of the item tags involved in the above embodiments are as follows Sk1 to Sk3:

[0532] Sk1. The target object proposes a guide requirement, and searches for item tags of relevant locations in the favorites;

[0533] Sk2. Provide suggestions on relevant items for the target object in the form of tips outside the itinerary, such as special products that can be purchased;

[0534] Sk3. Search for shops where the item can be purchased through the Internet, find shops within 1 km of the itinerary route, and plan the shop locations in the itinerary plan.

[0535] In an embodiment of the present application, if the travel guide requirements proposed by the target object include specific time information, search for events, scenic spots, and items related to the time tag in the favorite content; extract the event, scenic spot, and item information that meets the time tag, and filter out unsuitable scenic spots, events, and items in the itinerary for the target object according to the time information, and generate relevant suggestions; and obtain additional information through an online search and provide it to the target object.

[0536] Specifically, the usage scenarios of the time tags involved in the above embodiments are as follows Sl1 to Sl4:

[0537] Sl1. When the requirements of the target object include specific time information, search for events, scenic spots, and items related to the time tag in the favorites;

[0538] Sl2. Extract the event, scenic spot, and item information that meets the time tag, and filter out unsuitable scenic spots, events, and items in the itinerary for the target object according to the time information;

[0539] Sl3. Provide additional suggestions outside the itinerary for the target object in the form of tips. For example, it is not suitable to go to location H in city X in March;

[0540] Sl4. It is also possible to obtain additional information through an online search, such as the weather corresponding to the date, and provide it to the target object in the form of tips.

[0541] (2) Word query scenario.

[0542] The specific summarization process is as follows, Sm1 to Sm5:

[0543] Sm1. Management of favorite content: Establish a management system for favorite content in the APP, and label each piece of content according to the difficulty tags generated during favoriting;

[0544] Sm2. Retrieval of favorite content: Retrieve the content favorited by the target object from the stored data;

[0545] Sm3. AI summarizes the historical content of the target object: The AI obtains all the words, phrases, and sentences queried by the target object today based on the browsing record, and through the certainty of the existing database, assigns a difficulty score to each historical sentence;

[0546] Sm4. Label the historical content and favorite content in two dimensions: Content type dimension: words, phrases, sentences; Content difficulty dimension: one star to five stars;

[0547] Sm5. The target object can filter the summarized knowledge points through two dimensions.

[0548] It should be noted that the above-listed summarization methods are only simple examples. In addition, other summarization methods are also applicable to the embodiments of this application, and will not be elaborated one by one here.

[0549] Similar to the client side above, it should be emphasized that the relevant data related to the object's multiple searches during information search in the above-mentioned server-side technical solutions. The collection, use, and processing of this data need to comply with the relevant laws, regulations, and standards of relevant countries and regions, and the object's permission or consent has been obtained, and will not be repeated here.

[0550] Based on the same inventive concept, the embodiments of this application also provide an information search device. As Figure 28 shown, it is a schematic structural diagram of the information search device 2800, and may include:

[0551] The first response unit 2801 is configured to present first search suggestion information in the information search interface in response to a search operation triggered by the target object based on the target search information, where the first search suggestion information is the summary information of the current search result determined to meet the preset presentation condition, and the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object;

[0552] The second response unit 2802 is configured to present a search result interface corresponding to the target search information in response to a viewing operation triggered based on the first search suggestion information. The current search result in the search result interface is generated based on the target search information and the context corpus information of the target object. The context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

[0553] Optionally, when the target search information is related to a geographical location, the search result interface further includes a map module. The map module is configured to present: the favorite locations related to the geographical location in the favorite content of the target object, the recommended locations related to the target search information determined based on a preset recommendation rule, and the driving route between the favorite location and the recommended location.

[0554] The current search result in the search result interface includes the basic information of each recommended location and at least one web page result associated with the recommended location.

[0555] Optionally, the second response unit 2802 is further configured to:

[0556] Update the search information in the search information input area of the search result interface from the target search information to the combination of the target search information and the number of target favorite contents. The target favorite content is the content in the favorite content of the target object that is related to the current search operation; or

[0557] Update the search information in the search information input area of the search result interface from the target search information to the first search suggestion information.

[0558] Optionally, the search result interface further includes additional prompt information that matches the current search result. The additional prompt information is: prompt information related to at least one of the date information and the location information in the current search process.

[0559] Optionally, the search result interface further includes an intelligent Q&A entry. The apparatus further includes:

[0560] The third response unit 2803 is configured to switch from the network search mode to the dialogue search mode in response to a dialogue operation triggered based on the intelligent Q&A entry.

[0561] In response to a question operation for the current search result in the dialogue search mode, present the first question input for the current search result and the first answer information corresponding to the first question.

[0562] Optionally, the search result interface further includes at least one answer operation control. Then the apparatus further includes:

[0563] The fourth response unit 2804 is configured to, in response to a selection operation on a target answer operation control among at least one answer operation control, perform an operation corresponding to the target answer operation control on the first answer information, and present a corresponding operation execution result.

[0564] Optionally, the apparatus further includes:

[0565] The fifth response unit 2805 is configured to, in response to a content selection operation triggered based on a search result interface, identify the target content selected this time in the search result interface, and the intelligent recognition information corresponding to the target content;

[0566] Wherein, the intelligent recognition information includes at least one of the following:

[0567] Key information included in the target content, the type to which the target content belongs, and the importance level of the target content.

[0568] Optionally, the fifth response unit 2805 is further configured to:

[0569] In response to a question operation on the target content, present a corresponding answer interface, where the answer interface includes the target content, a second question input for the target content, and second answer information corresponding to the second question.

[0570] Optionally, the answer interface further includes an intelligent Q&A entry; the fifth response unit 2805 is further configured to:

[0571] In response to a conversation operation triggered based on the intelligent Q&A entry, switch from the network search mode to the conversation search mode;

[0572] In response to a question operation on the second answer information in the conversation search mode, present a third question input for the second answer information, and third answer information corresponding to the third question.

[0573] Optionally, the fifth response unit 2805 is further configured to:

[0574] Collect the target content, and present a corresponding collection animation in the search result interface; wherein, the collection animation indicates that the target content flies into the collection control in the search result interface.

[0575] Optionally, the apparatus further includes:

[0576] The sixth response unit 2806 is configured to, in response to a viewing operation on the collected content of a target object, present a corresponding collection interface, and in the collection interface, label different classification tags with different marking styles;

[0577] Wherein, for repeated classification tags, the number of repetitions is marked at the relevant positions.

[0578] Optionally, the device further includes:

[0579] A summary unit 2807, configured to present at least one second search suggestion message in the information search interface in response to a summary operation triggered by a target object. Each second search suggestion message corresponds to a search topic, and the second search suggestion message is a summary message corresponding to the search topic. The search topic is obtained by dividing multiple search messages related to the target object;

[0580] In response to a selection operation on a target search suggestion message among at least one second search suggestion message, present a search result summary interface corresponding to the target search suggestion message. The search result summary interface includes summary content of search results corresponding to each search message under the corresponding search topic.

[0581] Optionally, if the search topic is a travel guide, the summary content is a travel guide including a travel plan corresponding to at least one travel time period, and the favorite content referred to by the travel guide is marked;

[0582] If the search topic is knowledge point learning, the summary content includes each knowledge point divided according to importance degree, and a knowledge example integrating each knowledge point.

[0583] Based on the same inventive concept, an embodiment of the present application further provides another information search device. As Figure 29 shown, it is a schematic structural diagram of an information search device 2900, and may include:

[0584] An information acquisition unit 2901, configured to acquire context corpus information of a target object. The context corpus information includes at least one of the following: search information of each search operation of the target object, content browsed by the target object in each search result, and content of feedback behavior implemented by the target object;

[0585] A result generation unit 2902, configured to, after receiving a search request triggered by target search information, if it is determined that a preset presentation condition is satisfied, input the target search information and the context corpus information into a target language model to generate a summary message of the current search result and the current search result. The preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object;

[0586] A feedback unit 2903, configured to feedback the summary message and the current search result to the client, so that after the client determines that the preset presentation condition is satisfied, present a first search suggestion message including the summary message in the information search interface, and present a search result interface including the current search result after responding to a view operation triggered by the first search suggestion message.

[0587] Optionally, if the target search information is related to the geographical location, the result generation unit 2902 is specifically configured to:

[0588] Determine the collection locations related to the geographical location in the collection content of the target object;

[0589] Input the collection location into the target language model, use the target language model to learn the feature information of the collection location, and combine the historical behavior of the target object to determine at least one recommended location for the target object;

[0590] For each recommended location, by collecting the content related to the recommended location in the network, extract the basic information of the recommended location and at least one associated web page result, and use them as the search results for this time.

[0591] Optionally, the search result interface further includes additional prompt information matching the search results for this time; the additional prompt information is: prompt information related to at least one of the date information and location information in the search process for this time; the result generation unit 2902 is further configured to determine the additional prompt information in the following manner and then feedback it to the client through the feedback unit 2903:

[0592] Extract the date information and location information in the search process for this time;

[0593] Based on the date information and location information, generate the prompt information corresponding to each preset prompt template;

[0594] Sort the prompt information corresponding to each prompt template according to their respective preset weights;

[0595] Use the prompt information within the specified order range of the sorting result as the additional prompt information.

[0596] Optionally, if the target search information is related to knowledge point learning and the target language model is a knowledge point recognition model, the result generation unit 2902 is specifically configured to:

[0597] Input the knowledge points to be searched corresponding to the target search information into the trained knowledge point recognition model;

[0598] Based on the knowledge point recognition model, if it is determined that the associated knowledge points of the knowledge points to be searched have been searched by the target object, then use the knowledge points to be searched, the determined associated knowledge points of the knowledge points to be searched, and the relevant explanations as the search results for this time.

[0599] Optionally, the device further includes:

[0600] An intelligent processing unit 2904, configured to determine the selected target content according to the selection range of the target object after the target object makes a content selection for the search results for this time;

[0601] After detecting a voice input event, convert the voice of the question input by the target object for the target content into text;

[0602] Perform semantic analysis and keyword extraction on the text to obtain a text recognition result;

[0603] Input the text recognition result and the target content into the target language model to obtain answer information generated by the target language model for the target content, and feedback it to the client, so that the client presents a corresponding answer interface in response to a question operation for the target content. The answer interface includes the target content, the question, and the answer information.

[0604] Optionally, the device further includes:

[0605] A summary unit 2905, configured to respectively determine search topics corresponding to multiple search information related to the target object after receiving a summary request triggered by the target object;

[0606] Through the target language model, summarize the search results of different search topics in units of search topics to obtain summary content corresponding to each search topic;

[0607] Score and sort the obtained summary content, and filter out the summary content with a score lower than a preset score threshold;

[0608] Generate second search suggestion information corresponding to the remaining summary content respectively, and feedback the generated second search suggestion information to the client, so that the client presents at least one second search suggestion information in the information search interface in response to a summary operation, and presents a search result summary interface corresponding to the target search suggestion information after a selection operation on the target search suggestion information in at least one second search suggestion information. The search result summary interface includes summary content under the corresponding search topic; each second search suggestion information is summary information corresponding to the corresponding search topic.

[0609] Optionally, the device further includes:

[0610] A collection unit 2906, configured to classify the collection content of the target object according to a preset dimension, and label classification tags included in each collection content; the collection content of the target object includes at least one of the following: content browsed by the target object in each search result, content for which the target object performs a feedback behavior;

[0611] The summary unit 2905 is specifically configured to:

[0612] Retrieve target collection content related to the target search information from the collection content according to the classification tags;

[0613] Through the target language model, taking the search topic as a unit, combining the target collection content, the search results of different search topics are summarized to obtain the summary content corresponding to each search topic.

[0614] Since this application is based on the network search experience, every time the object conducts a search, the search information input by the object, the content browsed by the object in each search result, and the content of the feedback behavior implemented by the object are all statistically analyzed and constructed into the context corpus information of this object. Furthermore, when the object conducts a search based on the target search information, first analyze whether this search operation is related to the previous historical search operations. If it is related, then the context corpus information of the object and the target search information input by the object this time can be combined and input into the target language model. By utilizing the capabilities of the language model, the present search result after multiple rounds of information processing that conforms to the personal preferences of the object can be provided for the object.

[0615] Moreover, when it is determined that this search operation is related to the previous historical search operations, the first search suggestion information is dynamically presented in the information search interface, prompting the object that this search is related to the previous search and prompting the summary information of the present search result of the object, so as to present the present search result generated after multiple rounds of information processing after it is determined that the object clicks to view the first search suggestion information. In this way, even for some relatively complex search problems, it is not necessary for the object to conduct multiple cross-platform searches by itself, but the answer that meets the needs of the object is automatically generated in combination with the personalized corpus information of the object, effectively improving the efficiency of the object's information search.

[0616] For the convenience of description, the above parts are divided into each module (or unit) according to functions and described separately. Of course, when implementing this application, the functions of each module (or unit) can be implemented in the same or multiple software or hardware.

[0617] In the embodiments of this application, the term "module" or "unit" refers to a computer program with a predetermined function or a part of a computer program, which works together with other related parts to achieve a predetermined goal, and can be fully or partially implemented by using software, hardware (such as a processing circuit or a memory), or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be a part of the overall module or unit that includes the function of this module or unit.

[0618] After introducing the information search method and device of the exemplary embodiment of this application, next, an electronic device according to another exemplary embodiment of this application is introduced.

[0619] Those skilled in the art can understand that various aspects of the present application can be implemented as a system, method, or program product. Therefore, various aspects of the present application can be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which can be collectively referred to as "circuit", "module", or "system" here.

[0620] Based on the same inventive concept as the above method embodiment, an electronic device is also provided in an embodiment of the present application. In one embodiment, the electronic device can be a server, such as Figure 1 the server 120 shown. In this embodiment, the structure of the electronic device can be as Figure 30 shown, including a memory 3001, a communication module 3003, and one or more processors 3002.

[0621] The memory 3001 is used to store the computer program executed by the processor 3002. The memory 3001 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system and programs required to run the instant messaging function, etc.; the data storage area can store various instant messaging information and operation instruction sets, etc.

[0622] The memory 3001 can be a volatile memory, such as a random-access memory (RAM); the memory 3001 can also be a non-volatile memory, such as a read-only memory, a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); or the memory 3001 is any other medium that can be used to carry or store the desired computer program in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 3001 can be a combination of the above memories.

[0623] The processor 3002 can include one or more central processing units (CPUs) or be a digital processing unit, etc. The processor 3002 is used to implement the above information search method when calling the computer program stored in the memory 3001.

[0624] The communication module 3003 is used to communicate with terminal devices and other servers.

[0625] In the embodiment of the present application, the specific connection medium between the above memory 3001, communication module 3003, and processor 3002 is not limited. The embodiment of the present application is inFigure 30 In the [device], a memory 3001 and a processor 3002 are connected through a bus 3004. The bus 3004 is described by a thick line in the [device]. The connection manners between other components are only for illustrative purposes and are not limited thereto. The bus 3004 can be divided into an address bus, a data bus, a control bus, etc. For the convenience of description, Figure 30 in the [device], it is only described by a thick line, but it does not describe that there is only one bus or one type of bus. Figure 30

[0626] The memory 3001 stores a computer storage medium, and the computer storage medium stores computer-executable instructions for implementing the information search method of the embodiments of the present application. The processor 3002 is configured to execute the above-mentioned information search method, as Figure 26 shown.

[0627] In another embodiment, the electronic device can also be other electronic devices, such as Figure 1 the terminal device 110 shown. In this embodiment, the structure of the electronic device can be as Figure 31 shown, including: a communication component 3110, a memory 3120, a display unit 3130, a camera 3140, a sensor 3150, an audio circuit 3160, a Bluetooth module 3170, a processor 3180 and other components.

[0628] The communication component 3110 is configured to communicate with a server. In some embodiments, it may include a Wireless Fidelity (WiFi) module. The WiFi module belongs to short-range wireless transmission technology, and the electronic device can help users send and receive information through the WiFi module.

[0629] The memory 3120 can be used to store software programs and data. The processor 3180 executes various functions and data processing of the terminal device 110 by running the software programs or data stored in the memory 3120. The memory 3120 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices. The memory 3120 stores an operating system that enables the terminal device 110 to operate. In the present application, the memory 3120 can store an operating system and various application programs, and can also store a computer program for executing the information search method of the embodiments of the present application.

[0630] ​The display unit 3130 can also be used to display the information input by the user or the information provided to the user, as well as the graphical user interface (GUI) of various menus of the terminal device 110. Specifically, the display unit 3130 may include a display screen 3132 disposed on the front of the terminal device 110. Among them, the display screen 3132 can be configured in the form of a liquid crystal display, a light-emitting diode, etc. The display unit 3130 can be used to display the information search interface, search result interface, answer interface, etc. in the embodiments of the present application.

[0631] The display unit 3130 can also be used to receive the input digital or character information and generate signal inputs related to the user settings and function control of the terminal device 110. Specifically, the display unit 3130 may include a touch screen 3131 disposed on the front of the terminal device 110, which can collect the touch operations of the user on or near it, such as clicking buttons, dragging scroll boxes, etc.

[0632] Among them, the touch screen 3131 can cover the display screen 3132, or the touch screen 3131 and the display screen 3132 can be integrated to implement the input and output functions of the terminal device 110. After integration, it can be simply called a touch display screen. In the present application, the display unit 3130 can display application programs and corresponding operation steps.

[0633] The camera 3140 can be used to capture static images, and the user can publish the images captured by the camera 3140 through an application. The camera 3140 can be one or multiple. The object generates an optical image through the lens and projects it onto the photosensitive element. The photosensitive element can be a charge coupled device (CCD) or a complementary metal-oxide-semiconductor (CMOS) phototransistor. The photosensitive element converts the optical signal into an electrical signal, and then transmits the electrical signal to the processor 3180 to convert it into a digital image signal.

[0634] The terminal device may further include at least one sensor 3150, such as an acceleration sensor 3151, a distance sensor 3152, a fingerprint sensor 3153, a temperature sensor 3154. The terminal device may also be configured with other sensors such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, a light sensor, a motion sensor, etc.

[0635] The audio circuit 3160, the speaker 3161, and the microphone 3162 can provide an audio interface between the user and the terminal device 110. The audio circuit 3160 can transmit the electrical signal converted from the received audio data to the speaker 3161, and the speaker 3161 converts it into a sound signal for output. The terminal device 110 can also be configured with volume buttons for adjusting the volume of the sound signal. On the other hand, the microphone 3162 converts the collected sound signal into an electrical signal, which is received by the audio circuit 3160, converted into audio data, and then the audio data is output to the communication component 3110 to be sent to, for example, another terminal device 110, or the audio data is output to the memory 3120 for further processing.

[0636] The Bluetooth module 3170 is used to interact with other Bluetooth devices having Bluetooth modules through the Bluetooth protocol. For example, the terminal device can establish a Bluetooth connection with a wearable electronic device (such as a smart watch) that also has a Bluetooth module through the Bluetooth module 3170 to perform data interaction.

[0637] The processor 3180 is the control center of the terminal device, connecting various parts of the entire terminal using various interfaces and lines. By running or executing software programs stored in the memory 3120 and calling data stored in the memory 3120, it executes various functions of the terminal device and processes data. In some embodiments, the processor 3180 may include one or more processing units; the processor 3180 can also integrate an application processor and a baseband processor. Among them, the application processor mainly processes the operating system, user interface, and application programs, etc., and the baseband processor mainly processes wireless communication. It can be understood that the above baseband processor may not be integrated into the processor 3180. In this application, the processor 3180 can run the operating system, application programs, user interface display, and touch response, as well as the information search method of the embodiments of this application. In addition, the processor 3180 is coupled to the display unit 3130.

[0638] In some possible implementation manners, various aspects of the information search method provided in this application can also be implemented in the form of a program product, which includes a computer program. When the program product runs on an electronic device, the computer program is used to cause the electronic device to execute the steps in the information search method according to various exemplary embodiments of this application described above in this specification. For example, the electronic device can execute steps such as Figure 2 or Figure 26 shown in.

[0639] The program product may employ any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0640] The program product of the embodiments of the present application may employ a portable compact disc read-only memory (CD-ROM) and include a computer program, and may be run on an electronic device. However, the program product of the present application is not limited thereto. In this document, the readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0641] The readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a readable computer program. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the foregoing. The readable signal medium may also be any readable medium other than the readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0642] The computer program contained on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0643] The computer program for performing the operations of the present application may be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The computer program may be executed entirely on the user's electronic device, partially on the user's electronic device, executed as a stand-alone software package, partially on the user's electronic device and partially on a remote electronic device, or entirely on a remote electronic device or server. In the case of a remote electronic device, the remote electronic device may be connected to the user's electronic device through any type of network including a local area network (LAN) or a wide area network (WAN), or may be connected to an external electronic device (e.g., through the Internet using an Internet service provider).

[0644] It should be noted that although several units or subunits of the device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0645] In addition, although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution.

[0646] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) that contain computer-usable computer programs.

[0647] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0648] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that realizes the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0649] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are executed on the computer or other programmable apparatus to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one process or a plurality of processes and / or blocks Figure 1 one process or a plurality of processes and / or blocks Figure 1 steps for implementing the functions specified in one block or a plurality of blocks.

[0650] Although the preferred embodiments of the present application have been described, additional changes and modifications can be made to these embodiments by those skilled in the art once they learn of the basic creative concept. Therefore, the appended claims are intended to be construed to cover the preferred embodiments as well as all changes and modifications that fall within the scope of the present application.

[0651] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. An information search method, characterized in that, The method includes: In response to a search operation triggered by a target object based on target search information, presenting first search suggestion information in an information search interface, where the first search suggestion information is a summary information of the current search results determined to meet a preset presentation condition, and the preset presentation condition is that the current search operation of the target object is related to the historical search operations of the target object; In response to a viewing operation triggered based on the first search suggestion information, presenting a search result interface corresponding to the target search information, where the current search results in the search result interface are generated based on the target search information and the context corpus information of the target object, and the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

2. The method according to claim 1, characterized in that When the target search information is related to a geographical location, the search result interface further includes a map module; the map module is used to present: the favorite locations related to the geographical location in the favorite content of the target object, the recommended locations related to the target search information determined based on a preset recommendation rule, and the driving route between the favorite locations and the recommended locations; The current search results in the search result interface include the basic information of each recommended location and at least one web page result associated with the recommended location.

3. The method according to claim 1, wherein The presenting the search result interface corresponding to the target search information in response to a viewing operation triggered based on the first search suggestion information further includes: Updating the search information in the search information input area of the search result interface from the target search information to a combination of the target search information and the number of target favorite contents; the target favorite contents are the contents in the favorite content of the target object that are related to the current search operation; or Updating the search information in the search information input area of the search result interface from the target search information to the first search suggestion information.

4. The method according to claim 1, wherein The search result interface further includes additional prompt information that matches the current search results; where the additional prompt information is: prompt information related to at least one of the date information and the location information in the current search process.

5. The method according to claim 1, wherein The search result interface further includes a smart Q&A entry; the method further includes: In response to a conversation operation triggered based on the smart Q&A entry, switching from a web search mode to a conversation search mode; In response to a question operation for the current search results in the conversation search mode, presenting a first question input for the current search results and first answer information corresponding to the first question.

6. The method according to claim 5, wherein The search result interface further includes at least one answer operation control; then the method further includes: In response to a selection operation on a target answer operation control among the at least one answer operation control, performing an operation corresponding to the target answer operation control on the first answer information and presenting a corresponding operation execution result.

7. The method according to any one of claims 1 to 6, characterized in that The method further includes: In response to a content selection operation triggered based on the search result interface, identify the target content selected this time in the search result interface, as well as the intelligent recognition information corresponding to the target content; Among them, the intelligent recognition information includes at least one of the following: The key information contained in the target content, the type to which the target content belongs, the importance level of the target content.

8. The method according to claim 7, wherein The method further includes: In response to a question operation for the target content, present a corresponding answer interface, where the answer interface includes the target content, a second question input for the target content, and second answer information corresponding to the second question.

9. The method according to claim 8, wherein The answer interface further includes an intelligent Q&A entry; the method further includes: In response to a conversation operation triggered based on the intelligent Q&A entry, switch from the network search mode to the conversation search mode; In response to a question operation for the second answer information in the conversation search mode, present a third question input for the second answer information, and third answer information corresponding to the third question.

10. The method according to claim 7, characterized in that, The method further includes: Collect the target content and present a corresponding collection animation in the search result interface; where the collection animation indicates that the target content flies into the collection control in the search result interface.

11. The method according to any one of claims 1 to 6, characterized in that, The method further includes: In response to a viewing operation for the collected content of the target object, present a corresponding collection interface, and in the collection interface, mark different classification labels with different annotation styles; Among them, for the repeatedly occurring classification labels, mark the number of repeated occurrences at the relevant positions.

12. The method according to any one of claims 1 to 6, characterized in that The method further includes: In response to a summarization operation triggered by the target object, present at least one second search suggestion information in the information search interface, each second search suggestion information corresponds to a search topic, and the second search suggestion information is the abstract information corresponding to the search topic; the search topic is obtained by dividing multiple search information related to the target object; In response to a selection operation for the target search suggestion information in the at least one second search suggestion information, present a search result summary interface corresponding to the target search suggestion information, where the search result summary interface includes the summary content of the search results corresponding to each search information under the corresponding search topic.

13. The method according to claim 12, characterized in that, If the search topic is a travel guide, the summary content is a travel guide including a travel plan corresponding to at least one travel time period, and the collected content referred to by the travel guide is marked; If the search topic is knowledge point learning, the summary content includes each knowledge point divided according to the importance level, and a knowledge example integrating each knowledge point.

14. An information search method, characterized in that, The method includes: Obtain the context corpus information of the target object, where the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, the content of the feedback behavior implemented by the target object; After receiving a search request triggered by target search information, if it is determined that the preset presentation condition is met, the target search information and the context corpus information are input into the target language model to generate the current search result and the summary information of the current search result; the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object; The summary information and the current search result are fed back to the client, so that after the client determines that the preset presentation condition is met, the first search suggestion information including the summary information is presented in the information search interface, and after responding to the view operation triggered by the first search suggestion information, the search result interface including the current search result is presented.

15. The method according to claim 14, wherein If the target search information is related to a geographical location, the target search information and the context corpus information are input into the target language model, and the corresponding current search result is generated based on the target language model, including: Determine the collection locations related to the geographical location in the collection content of the target object; Input the collection location into the target language model, use the target language model to learn the feature information of the collection location, and combine the historical behavior of the target object to determine at least one recommended location for the target object; For each recommended location, by collecting the content related to the recommended location in the network, extract the basic information of the recommended location and at least one associated web page result, and use them as the current search result.

16. The method according to claim 14, wherein The search result interface further includes additional prompt information matching the current search result; the additional prompt information is: prompt information related to at least one of the date information and the location information in the current search process; the additional prompt information is fed back to the client after being determined by the following method: Extract the date information and the location information in the current search process; Generate the prompt information corresponding to each preset prompt template based on the date information and the location information; Sort the prompt information corresponding to each prompt template according to their respective preset weights; Use the prompt information within the specified order range of the sorting result as the additional prompt information.

17. The method according to claim 14, characterized in that If the target search information is related to knowledge point learning and the target language model is a knowledge point recognition model, the target search information and the context corpus information are input into the target language model, and the corresponding current search result is generated based on the target language model, including: Input the knowledge point to be searched corresponding to the target search information into the trained knowledge point recognition model; Based on the knowledge point recognition model, if it is determined that the associated knowledge point of the knowledge point to be searched has been searched by the target object, use the knowledge point to be searched, the determined associated knowledge point of the knowledge point to be searched, and the relevant explanations as the current search result.

18. The method according to claim 14, wherein The method further includes: After the target object selects content from the current search result, determine the selected target content according to the selection range of the target object; After detecting a voice input event, convert the voice of the question input by the target object for the target content into text; Perform semantic analysis and keyword extraction on the text to obtain a text recognition result; Input the text recognition result and the target content into the target language model to obtain answer information generated by the target language model for the target content, and feedback it to the client, so that the client presents a corresponding answer interface in response to a question operation for the target content, and the answer interface includes the target content, the question, and the answer information.

19. The method according to any one of claims 14 to 18, characterized in that The method further includes: After receiving a summary request triggered by the target object, respectively determine the search topics corresponding to multiple search information related to the target object; Through the target language model, summarize the search results of different search topics in units of search topics to obtain summary content corresponding to each search topic; Score and rank the obtained summary content, and filter out the summary content with a score lower than a preset score threshold; Generate second search suggestion information corresponding to the remaining summary content respectively, and feedback the generated second search suggestion information to the client, so that the client presents at least one piece of the second search suggestion information in the information search interface in response to a summary operation, and presents a search result summary interface corresponding to the target search suggestion information after responding to a selection operation on the target search suggestion information in the at least one piece of second search suggestion information, and the search result summary interface includes the summary content under the corresponding search topic; each piece of second search suggestion information is the abstract information corresponding to the corresponding search topic.

20. The method according to claim 19, wherein The method further includes: Classify the favorite content of the target object according to a preset dimension, and label the classification tags included in each piece of favorite content; the favorite content of the target object includes at least one of the following: the content browsed by the target object in each search result, the content for which the target object performs a positive feedback behavior; The step of summarizing the search results of different search topics in units of search topics through the target language model to obtain summary content corresponding to each search topic includes: Retrieve the target favorite content related to the target search information from the favorite content according to the classification tag; Through the target language model, summarize the search results of different search topics in units of search topics in combination with the target favorite content to obtain summary content corresponding to each search topic.

21. An information search device, characterized in that, It includes: A first response unit, configured to present first search suggestion information in an information search interface in response to a search operation triggered by a target object based on target search information, where the first search suggestion information is the abstract information of the current search results presented by determining that a preset presentation condition is met, and the preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object; A second response unit, configured to present a search result interface corresponding to the target search information in response to a viewing operation triggered based on the first search suggestion information, where the current search result in the search result interface is generated based on the target search information and the context corpus information of the target object, and the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object.

22. An information search device, characterized in that, It includes: An information acquisition unit, configured to acquire the context corpus information of the target object, where the context corpus information includes at least one of the following: the search information of each search operation of the target object, the content browsed by the target object in each search result, and the content of the feedback behavior implemented by the target object; A result generation unit, configured to, after receiving a search request triggered based on the target search information, if it is determined that a preset presentation condition is met, input the target search information and the context corpus information into a target language model to generate the current search result and the abstract information of the current search result; The preset presentation condition is that the current search operation of the target object is related to the historical search operation of the target object; A feedback unit, configured to feedback the abstract information and the current search result to the client, so that after the client determines that the preset presentation condition is met, it presents the first search suggestion information including the abstract information in the information search interface, and presents a search result interface including the current search result in response to a viewing operation triggered based on the first search suggestion information.

23. An electronic device, characterized in that, It includes a processor and a memory, where the memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to execute the steps of any one of claims 1 to 20.

24. A computer-readable storage medium, characterized in that, It includes a computer program, and when the computer program runs on an electronic device, the computer program is used to cause the electronic device to execute the steps of any one of claims 1 to 20.

25. A computer program product, characterized in that, It includes a computer program, and the computer program is stored in a computer-readable storage medium; when the processor of the electronic device reads the computer program from the computer-readable storage medium, the processor executes the computer program, so that the electronic device executes the steps of any one of claims 1 to 20.

Images