Information search method and device, storage medium and electronic equipment

Abstract

The invention provides an information search method, an information search device, a computer storage medium and electronic equipment, and relates to the technical field of information processing. Comprising the steps that an information search request is received, and the information search request at least comprises input to-be-searched information; determining an information category corresponding to the to-be-searched information; in response to the fact that the information category corresponding to the to-be-searched information contains a first information category, calling a local search engine to search to obtain a first search result; in response to the fact that the information category corresponding to the to-be-searched information contains a second information category, calling the external search engine to search to obtain a second search result; and determining a first target search result corresponding to the to-be-searched information according to the first search result and the second search result. The information search and information aggregation capabilities of the system can be improved, the universality and accuracy of the information search result are improved, and then the experience feeling of user information acquisition is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of information processing technology, and in particular to an information search method, an information search device, a computer storage medium, and an electronic device. Background Art

[0002] With the rapid development and widespread application of artificial intelligence and large models, intelligent search based on large models (Large Model), also known as foundation models, can assist users in obtaining information more efficiently with its powerful semantic understanding, information aggregation and continuous learning capabilities, thus having a wide range of applications in corporate knowledge management, academic research, legal work, personal daily life and other fields.

[0003] However, at present, in the process of information search through large models, commercial search engine URLs are usually searched directly based on user input questions, which makes the search system's information search and information aggregation capabilities insufficient and highly limited, resulting in insufficient breadth and accuracy of information search results, affecting users' information search experience. Summary of the invention

[0004] The present disclosure provides an information search method, an information search device, a computer storage medium and an electronic device, thereby improving the information search and information aggregation capabilities of the system, improving the breadth and accuracy of information search results, and further improving the user's experience of obtaining information.

[0005] In a first aspect, an embodiment of the present disclosure provides an information search method, the method comprising: receiving an information search request, the information search request at least including input information to be searched; determining an information category corresponding to the information to be searched; in response to the information category corresponding to the information to be searched including a first information category, calling a local search engine to search and obtain a first search result; in response to the information category corresponding to the information to be searched including a second information category, calling an external search engine to search and obtain a second search result; and determining a first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0006] In an optional embodiment of the present disclosure, determining the information category corresponding to the information to be searched includes: preprocessing the information to be searched to obtain the preprocessed information to be searched; determining the information category corresponding to the preprocessed information to be searched; wherein the preprocessing method includes one or more of the following: converting the information format of the information to be searched; and expanding the information content of the information to be searched.

[0007] In an optional embodiment of the present disclosure, determining the information category corresponding to the information to be searched includes: preprocessing the information to be searched to obtain the preprocessed information to be searched; determining the information category corresponding to the preprocessed information to be searched;

[0008] The pretreatment method includes one or more of the following:

[0009] Convert the information format of the search information;

[0010] Expand the information content of the search information.

[0011] In an optional embodiment of the present disclosure, the external search engine at least includes a metadata search engine, and the metadata search engine includes multiple sub-search engines; in response to the information category corresponding to the information to be searched including the second information category, calling the external search engine to search and obtain the second search result, including: in response to the information category corresponding to the information to be searched including the second information category, calling the metadata search engine to send the information to be searched to multiple sub-search engines for searching based on the metadata search engine, and obtain search results corresponding to the multiple sub-search engines; integrating the search results corresponding to the multiple sub-search engines to obtain the second search result.

[0012] In an optional embodiment of the present disclosure, the method may further include: performing vectorization processing on the information to be searched in the first information category, or performing keyword extraction on the information to be searched in the first information category to obtain keywords corresponding to the information to be searched;

[0013] Among them, calling the local search engine to search and obtain the first search result includes: searching based on the similarity between the information to be searched after vectorization processing and the search contents in the local search engine, or searching based on the similarity between the keywords corresponding to the information to be searched and the search contents in the local search engine to obtain the first search result.

[0014] In an optional embodiment of the present disclosure, the method may further include: constructing the first target search result with prompt words, and inputting the search result after the prompt word construction into a large language LLM model to generate a second target search result.

[0015] In an optional embodiment of the present disclosure, the method may further include: associating the information to be searched with the second target search results to generate first citation information; associating each search result in the second target search results with the corresponding search engine to generate second citation information; generating citation information based on the first citation information and / or the second citation information so as to synchronously display the second search results and the corresponding citation information.

[0016] In an optional embodiment of the present disclosure, the method may further include: in response to the information category corresponding to the information to be searched being the third information category, inputting the information to be searched into the large language model (LLM) to obtain a third search result.

[0017] In a second aspect, an embodiment of the present disclosure provides an information search device, which includes: a request receiving module, used to receive an information search request, the information search request at least including input information to be searched; a category confirmation module, used to determine the information category corresponding to the information to be searched; a first search module, used to call a local search engine to search and obtain a first search result in response to the information category corresponding to the information to be searched including a first information category; a second search module, used to call an external search engine to search and obtain a second search result in response to the information category corresponding to the information to be searched including a second information category; and a result determination module, used to determine a first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0018] In an optional embodiment of the present disclosure, the category confirmation module is used to preprocess the information to be searched to obtain the preprocessed information to be searched; determine the information category corresponding to the preprocessed information to be searched;

[0019] The pretreatment method includes one or more of the following:

[0020] Convert the information format of the search information;

[0021] Expand the information content of the search information.

[0022] In an optional embodiment of the present disclosure, the external search engine includes at least a metadata search engine, and the metadata search engine includes multiple sub-search engines; the second search module is used to call the metadata search engine in response to the information category corresponding to the information to be searched including the second information category, so as to send the information to be searched to the multiple sub-search engines for searching based on the metadata search engine, and obtain search results corresponding to the multiple sub-search engines; the search results corresponding to the multiple sub-search engines are integrated to obtain a second search result.

[0023] In an optional embodiment of the present disclosure, the device may also include a processing module, the processing module is used to perform vectorization processing on the information to be searched in the first information category, or to perform keyword extraction on the information to be searched in the first information category, to obtain keywords corresponding to the information to be searched; a first search module is used to perform a search based on the similarity between the information to be searched after vectorization processing and each search content in the local search engine, or based on the similarity between the keywords corresponding to the information to be searched and each search content in the local search engine, to obtain a first search result.

[0024] In an optional embodiment of the present disclosure, the device may further include an input module, which is used to construct a prompt word for the first target search result, and input the search result after the prompt word construction into the large language LLM model to generate a second target search result.

[0025] In an optional embodiment of the present disclosure, the processing module is also used to associate the information to be searched with the second target search results to generate first citation information; associate each search result in the second target search results with the corresponding search engine to generate second citation information; and generate citation information based on the first citation information and / or the second citation information so as to synchronously display the second search results and the corresponding citation information.

[0026] In an optional embodiment of the present disclosure, the input module is used for inputting the information to be searched into the large language model (LLM) to obtain a third search result in response to the information category corresponding to the information to be searched being the third information category.

[0027] In a third aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which implements the above information search method when executed by a processor.

[0028] In a fourth aspect, an embodiment of the present disclosure provides an electronic device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform the above information search method by executing the executable instructions.

[0029] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, including a computer program, which is executed by a processor to implement the above information search method.

[0030] The technical solution disclosed in this disclosure has the following beneficial effects:

[0031] The above-mentioned information search method receives an information search request, which at least includes input information to be searched; determines the information category corresponding to the information to be searched; in response to the information category corresponding to the information to be searched including the first information category, calls a local search engine to search and obtains a first search result; in response to the information category corresponding to the information to be searched including the second information category, calls an external search engine to search and obtains a second search result; and determines the first target search result corresponding to the information to be searched based on the first search result and the second search result. This method avoids the limitations of related technologies that use only one search engine for information retrieval by introducing multiple search engines to search for the information to be searched respectively, which not only expands the information source, improves the system's information search capability and information aggregation capability, but also improves the accuracy of the information retrieval results. On the other hand, this method also further deepens the understanding of the user's intention and returns more reasonable search results.

[0032] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A system architecture diagram showing one of the information search systems in this exemplary embodiment;

[0035] Figure 2 A flowchart showing one of the information search methods of this exemplary embodiment;

[0036] Figure 3 A flow chart showing a method for generating a second search result according to the exemplary embodiment is shown;

[0037] Figure 4 A schematic diagram showing one complete information search method according to the exemplary embodiment;

[0038] Figure 5 A schematic diagram of the structure of an information search device in this exemplary embodiment is shown;

[0039] Figure 6 A schematic structural diagram of an electronic device in this exemplary embodiment is shown. DETAILED DESCRIPTION

[0040] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, exemplary embodiments can be implemented in a variety of forms and should not be construed as being limited to the examples set forth herein; on the contrary, these embodiments are provided so that the present disclosure will be more comprehensive and complete, and the concepts of the exemplary embodiments are fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0041] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0042] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the steps. For example, some steps may be decomposed, while some steps may be combined or partially combined, so the actual execution order may change according to the actual situation.

[0043] Large Model, also known as Foundation Model, refers to a machine learning model with large-scale parameters and complex computing structure, which is used to improve the expressiveness and predictive performance of the model to handle more complex tasks and data. For example, natural language processing, computer vision, speech recognition and other tasks. At present, intelligent search based on large models is a new search mode. Compared with traditional search engines, search engines based on large models can be widely used in various fields with their powerful semantic understanding and generation capabilities, helping users to obtain information more efficiently. At the same time, it has the characteristics of strong semantic understanding ability, strong information aggregation ability, personalized recommendation, continuous learning, etc., and has a broad application prospect in the fields of enterprise knowledge management, academic research, legal work, personal daily learning, etc.

[0044] Among them, the Large Language Model (LLM) used in natural language processing is usually a natural language processing model with large-scale parameters and computing power. The LLM model can be trained with a large amount of text data and parameters to generate human-like natural language text or understand the meaning of language text. The LLM model is widely used in processing natural language tasks such as text generation, text classification, intelligent dialogue, question and answer, and is an important technical means to artificial intelligence.

[0045] The inventors found that in the process of information search in the intelligent search system based on the large model, the related technical solutions usually directly search for search engine URLs based on the search questions entered by the user in the search box, such as the search engines such as Bing, Baidu, Sougou, Google, etc., in order to obtain the corresponding answers to the questions. However, the above-mentioned related technical solutions make the information search and information aggregation capabilities of the intelligent search system low, which in turn affects the low breadth and accuracy of the search answers finally generated; and the system's semantic understanding ability and personalized recommendation ability are limited, and it is impossible to give full play to the advantages of the large model, which in turn affects the user's information acquisition experience.

[0046] In view of the above problems, the exemplary embodiments of the present disclosure propose an information search method, which introduces multiple search engines to search for the searched information respectively, avoiding the limitation of the related art that only one search engine is used for information retrieval, and not only expands the information source, but also improves the information search ability and information aggregation ability of the system, thereby improving the accuracy of the information retrieval results. On the other hand, the method also further deepens the understanding of the user's intention and returns more reasonable search results.

[0047] System architecture:

[0048] The following will be combined Figure 1 , an information search method and device proposed in the present disclosure are applied to Figure 1 In the system architecture of the exemplary application environment shown.

[0049] Figure 1 A system architecture diagram of an information search system according to the exemplary embodiment is shown as follows: Figure 1 As shown, the information search system 100 may include a terminal device 101 and a server 102 , and the terminal device 101 and the server 102 may be directly or indirectly connected to each other through one or more networks 103 .

[0050] Among them, the terminal device 101 can be, for example, a mobile phone, a tablet computer (PAD), a laptop computer, a desktop computer, a smart TV, a smart car device, a smart wearable device, a smart TV, and an aircraft, etc., any device that can perform any type of intelligent search. The terminal device 101 can be installed with a target application, and the target application can be an application program with functions such as inputting information to be searched and displaying search results. The application involved in the embodiment of the present disclosure can be a software client, or a client such as a web page or a small program. The server 102 is a server corresponding to the software or a web page, a small program, etc., and does not limit the specific type of the client.

[0051] The server 102 may be a background server of the target application, and is used to provide corresponding background services for it, such as providing search engine matching for the information to be searched, information search result generation services, etc. It may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides 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, i.e., content delivery networks (CDNs), as well as big data and artificial intelligence platforms, but is not limited thereto.

[0052] The network 103 may be a wired network or a wireless network. For example, the wireless network may be a mobile cellular network or a wireless fidelity (Wireless-Fidelity, WIFI) network. Of course, it may also be other possible networks, and the embodiments of the present disclosure do not impose any restrictions on this. It should be noted that, Figure 1 The figure is only an example. In fact, the number of terminal devices and servers is not limited and is not specifically limited in the embodiments of the present disclosure.

[0053] It should be noted that the information search method in the embodiment of the present disclosure can be executed by the terminal device 101 or the server 102 alone, or can be executed by the server 102 and the terminal device 101 together. When the information search method provided by the embodiment of the present disclosure is executed by the server 102 or the terminal device 101 alone, the above-mentioned application scenario may also only include a single device of the server 102 or the terminal device 101, or the server 102 and the terminal device 101 can also be considered to be the same device. Of course, in actual application, when the information search method provided by the embodiment of the present disclosure is executed by the server 102 and the terminal device 101 together, the server 102 and the terminal device 101 can also be the same device, that is, the server 102 and the terminal device 101 can be different functional modules of the same device, or virtual devices virtualized by the same physical device.

[0054] Among them, the server 102 and the terminal device 101 may include one or more processors, memories, and interactive I / O interfaces, etc. In addition, the server 102 may also be configured with a database, which may be used to store large model parameters or LLM model data, etc. Among them, the memories of the server 102 and the terminal device 101 may also store program instructions required to be executed in the information search method provided in the embodiment of the present disclosure, and these program instructions, when executed by the processor, can be used to implement the generation process of the information search results provided in the embodiment of the present disclosure.

[0055] For example, in an exemplary embodiment, a user can input information to be searched through an application such as smart question and answer and smart search running on the terminal device 101, and send a search request carrying the information to be searched to the server 102. The server 102 receives the search request information sent by the terminal device 101, and the search request at least includes the input information to be searched; the server 102 determines the information category corresponding to the information to be searched; in response to the information category corresponding to the information to be searched including the first information category, the local search engine is called to search and obtain the first search result; in response to the information category corresponding to the information to be searched including the second information category, the external search engine is called to search and obtain the second search result; the server 102 determines the first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0056] It should be noted that the above application scenarios are only shown to facilitate understanding of the spirit and principles of the present application, and the implementation methods of the present application are not limited in this regard.

[0057] Exemplary methods:

[0058] The following uses the server 102 as an execution subject and illustrates the information search method by applying it to the server 102. Figure 2 A flowchart showing one of the information search methods of this exemplary embodiment is shown. Figure 2 The information search method provided by the embodiment of the present disclosure includes the following steps 201 to 205:

[0059] Step 201: Receive an information search request, where the information search request at least includes input information to be searched.

[0060] Step 202: Determine the information category corresponding to the information to be searched.

[0061] Step 203: In response to the information category corresponding to the information to be searched including the first information category, a local search engine is called to perform a search to obtain a first search result.

[0062] Step 204: In response to the information category corresponding to the information to be searched including the second information category, an external search engine is called to perform a search to obtain a second search result.

[0063] Step 205: Determine a first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0064] pass Figure 2 In the embodiment shown, an information search request is received, wherein the information search request at least includes input information to be searched; the information category corresponding to the information to be searched is determined; in response to the information category corresponding to the information to be searched including a first information category, a local search engine is called to search and a first search result is obtained; in response to the information category corresponding to the information to be searched including a second information category, an external search engine is called to search and a second search result is obtained; based on the first search result and the second search result, a first target search result corresponding to the information to be searched is determined. This method avoids the limitations of related technologies that use only one search engine for information retrieval by introducing multiple search engines to search for the information to be searched respectively, which not only expands the information source, but also improves the information search capability and information aggregation capability of the system, thereby improving the accuracy of the information retrieval results. On the other hand, this method also further deepens the understanding of the user's intention and returns more reasonable search results.

[0065] The following will be combined with specific embodiments to Figure 2 The specific implementation of each step shown is exemplified.

[0066] In step 201, an information search request is received, where the information search request at least includes input information to be searched.

[0067] Among them, the information to be searched is the search content input by the user, which can be a question content, such as "How to search for information based on the big model?" "What is the big model?", or it can be a declarative sentence, such as "big model". The embodiments of the present disclosure do not impose any special restrictions on this.

[0068] Exemplarily, the user can input the information to be searched through the user interface provided by the terminal device and trigger an information search request to the server (eg, a WEB server). After receiving the information search request, the server can extract the information to be searched input by the user from the information search request.

[0069] It should be understood that the information to be searched contained in the information search request may be a combination of one or more of voice, text or pictures input by the user. The text information may be text directly input by the user or text converted by a speech-to-text model, and the embodiments of the present disclosure do not impose any special restrictions on this.

[0070] Step 202: Determine the information category corresponding to the information to be searched.

[0071] The information category in the embodiment of the present disclosure may be determined based on the application scenario and field involved in the content of the information to be searched, and may include a first information category, a second information category, and a third information category.

[0072] Among them, the application scenarios and fields involved in the first information category may be information categories related to internal enterprise information, which are not open to the external public domain. Therefore, it is usually impossible to obtain search results on the Internet. The application scenarios and fields involved in the second information category may be public domains, that is, information categories for which search results can be obtained on the Internet. For this, a search can be performed on the public Internet to obtain search results. The third information category belongs to relatively common and simple information categories, such as "Hello" and "Who are you".

[0073] Exemplarily, the information to be searched may be information input by a user, and after receiving the information to be searched input by the user, the server may directly determine the information category corresponding to the information to be searched.

[0074] In addition to the above-mentioned embodiment directly determining the information category of the information to be searched input by the user, in order to improve the search quality of the system, when executing step 202 to determine the information category corresponding to the information to be searched, in another optional embodiment of the present disclosure, the information to be searched is preprocessed to obtain the preprocessed information to be searched; and the search mode corresponding to the preprocessed information to be searched is determined.

[0075] In an optional embodiment, the pretreatment method includes one or more of the following:

[0076] Convert the information format of the search information;

[0077] Expand the information content of the search information.

[0078] Among them, for the above-mentioned information format conversion of the information to be searched, the information to be searched input by the user can be format converted to a format that is easier to search. For example, keywords are extracted from the information to be searched input by the user, and the information category is determined in the form of keywords.

[0079] Regarding the above-mentioned expansion of the information content of the information to be searched, when the user searches, the content of the information to be searched can be expanded and improved based on the context of the information to be searched, so as to expand the search scope, enrich the search results, and thus improve the ability and quality of information search.

[0080] In addition to the preprocessing method of converting the information format and expanding the information content of the information to be searched, other preprocessing methods may also be included, such as removing invalid information and data cleaning of the information to be searched. The embodiments of the present disclosure do not impose any special restrictions on this.

[0081] In step 203, in response to the information category corresponding to the information to be searched including the first information category, a local search engine is called to perform a search to obtain a first search result.

[0082] The application scenarios and fields involved in the first information category may be information related to internal enterprise information, which is not open to the external public domain. Therefore, it is usually impossible to obtain search results on the Internet. The local search engine (also called internal search engine) is a search engine built based on private data, and is usually used to retrieve internal private information.

[0083] For example, for the information to be searched in the first information category, since it is private information within the enterprise, the relevant technical solution is to directly call search engines such as Bing, Baidu, Sougou, and Google to search, but it is difficult to obtain matching search results.

[0084] In order to solve the problems existing in the above-mentioned related technologies, the embodiments of the present disclosure can call a local search engine to search the information to be searched when it is detected that the information to be searched is of the first information category, thereby obtaining a first search result.

[0085] When searching for information based on a local search engine, the search may be performed based on the following embodiments:

[0086] In an optional embodiment of the present disclosure, vectorization processing is performed on the information to be searched of the first information category to obtain the vectorized information to be searched.

[0087] Exemplarily, the information to be searched of the first information category can be vectorized based on the Embedding model, so that when the step of calling the local search engine to search and obtain the first search result is executed, the first search result is obtained by searching through the similarity between the vectorized information to be searched and the search contents in the local search engine.

[0088] Among them, the similarity between the information to be searched after vectorization processing and the search contents in the search engine can be determined based on cosine similarity or "dot product" distance. Based on the above similarity, not only can the search be implemented to improve the accuracy of the search results, but the search results can also be sorted to facilitate users to view the search results.

[0089] Alternatively, in another optional embodiment of the present disclosure, keywords are extracted from the information to be searched in the first information category to obtain keywords corresponding to the information to be searched, and a search is performed based on the similarity between the keywords corresponding to the information to be searched and each search content in the local search engine to obtain a first search result.

[0090] Exemplarily, a search may be performed in a local search engine using keywords, and when searching based on keywords, the search may be performed based on the similarity between the keywords and the search contents in the local search engine, thereby improving the accuracy of the search.

[0091] In an optional embodiment of the present disclosure, after a local search engine is called to search the information to be searched and a first search result is obtained, an event extraction operation may be performed on the first search result to generate a final search result.

[0092] Among them, the event extraction operation is to present the event information in the unstructured text into a structured form, which has important applications in the fields of public opinion monitoring, text summarization, automatic question answering, automatic construction of event graphs, etc. Event extraction itself includes two subtasks: event detection and argument detection. Event detection is to detect trigger words (words that best represent an event) and correctly determine its event type; argument detection is to detect the relevant elements of the event and correctly determine the role these elements play in the event.

[0093] Exemplarily, the event extraction operation can perform trigger word extraction, argument extraction, and attribute extraction on the first search result, so as to add different features to it and integrate it to form the final search result. This process can provide richness of the search results.

[0094] Step 204: In response to the information category corresponding to the information to be searched including the second information category, an external search engine is called to perform a search to obtain a second search result.

[0095] The application scenarios and fields involved in the second information category may be public domains, that is, information categories for which search results can be obtained on the Internet, and for which search results can be obtained on the public Internet.

[0096] External search engines are search engines that connect to external networks. They can be divided into paid search engines and metadata search engines. Among them, paid search engines such as Bing, Baidu, Sougou, Google, etc., use the APP_Key provided by the above search engines to call the interface to complete the search. For the above types of search engines, although they have good retrieval effects, stability and high efficiency, they need to be paid.

[0097] The metadata search engine is a non-paying search engine. It does not have a corresponding directory or database, but relies on multiple independent sub-search engines to provide search results. That is, when the server receives an information search request, the metadata search engine will distribute the information to be searched to multiple sub-search engines and integrate the results returned by each sub-search engine.

[0098] That is, in an optional embodiment of the present disclosure, the external search engine at least includes a metadata search engine, and the metadata search engine includes multiple sub-search engines; see Figure 3 In an optional embodiment of the present disclosure, in response to the information category corresponding to the information to be searched including the second information category, the above step 204 calls an external search engine to search and obtains the second search result, including the following steps 301-302:

[0099] Step 301: In response to the information category corresponding to the information to be searched including the second information category, a metadata search engine is called to send the information to be searched to multiple sub-search engines for searching based on the metadata search engine to obtain search results corresponding to the multiple sub-search engines.

[0100] Step 302: Integrate the search results corresponding to the multiple sub-search engines to obtain a second search result.

[0101] Among them, the multiple sub-search engines aggregated by the metadata search engine can be, for example, duckduckgo, yahoo, presearch, goo and other search engines, which are local private deployments. Compared with paid search engines, the search efficiency is lower and deployment and maintenance are required, but its search range is wide, the information sources are many, and it is free of charge.

[0102] In actual applications, users can flexibly choose to call metadata search engines or paid search engines for searching according to actual needs.

[0103] Exemplarily, the second search result obtained by calling an external search engine to search typically includes a web page title, a snapshot, and a uniform resource location (URL). All web pages in the second search result can be traversed to eliminate inaccessible web pages by detecting the accessibility of the URL, and then a hypertext markup language (html) parsing plug-in is used to obtain web page content, and data cleaning is performed on the web page content to eliminate invalid information in the web page. Finally, the web page summary, event content, etc. are extracted to obtain the final search result.

[0104] Step 205: Determine a first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0105] Exemplarily, after obtaining the first search result and the second search result, the first search result and the second search result may be integrated to obtain a first target search result corresponding to the information to be searched.

[0106] In addition to the above embodiment, the first target search result may be constructed with prompt words, and the search result after prompt word construction may be input into the large language LLM model to generate a second target search result.

[0107] Exemplarily, the LLM model is used for the integration and management of large model files, mainly providing functions such as downloading, storing, updating and deleting large models, and providing a variety of large model management services for large model reasoning engines. The LLM reasoning service provided by the LLM model can be used to understand the content in the search results, integrate the search content in the search results, write search results, and support distributed linear expansion and high-concurrency application scenarios.

[0108] In this embodiment, the first target search results obtained based on multiple search engines are combined with the large language model technology, so as to utilize the powerful semantic understanding and information analysis capabilities of the large language model to analyze, extract content, reorganize, generate answers and citations for the search results, and make the returned answers more logical, organized and personalized, thereby effectively improving the user search experience.

[0109] In addition, in an optional embodiment of the present disclosure, the information to be searched is associated with the second target search results to generate first citation information; each search result in the second target search results is associated with the corresponding search engine to generate second citation information; based on the first citation information and / or the second citation information, citation information is generated so as to synchronously display the second search results and the corresponding citation information.

[0110] For example, the information to be searched can be associated with the second target search results, and / or each search result in the second target search results can be associated with the corresponding search engine, and the generated citation information can help users understand its source and basis, and enhance credibility. Users can also further understand the search results of interest based on the citation information, thereby improving the user's information search experience.

[0111] Furthermore, in addition to any of the above embodiments, in an optional embodiment of the present disclosure, when the server detects that the information category corresponding to the information to be searched only contains the first information category, the server may call a local search engine to perform a search, obtain a first search result, and determine a first target search result corresponding to the information to be searched based on the first search result.

[0112] In this embodiment, the first target search result is the first search result, for which prompt words may be constructed, and the search result after prompt word construction is input into the large language LLM model to generate a second target search result.

[0113] In another optional embodiment of the present disclosure, when the server detects that the information category corresponding to the information to be searched only includes the second information category, the server may call an external search engine to search, obtain the second search result, and determine the first target search result corresponding to the information to be searched based on the second search result.

[0114] In this embodiment, the first target search result is the second search result. For this, the second search result can be constructed with prompt words, and the search result after the prompt word construction is input into the large language LLM model to generate the second target search result.

[0115] In another optional embodiment of the present disclosure, when the server detects that the information category corresponding to the information to be searched is the third information category, that is, in response to the information category corresponding to the information to be searched being the third information category, the server may input the information to be searched into the large language model (LLM) to obtain a third search result.

[0116] Among them, the information in the third information category is simple information, which does not need to be searched by calling a search engine, but only needs to be directly output by the LLM model.

[0117] Exemplarily, prompt words are first constructed for the information to be searched in the third information category, and then an inference service based on a large language model is performed to obtain a third search result, so as to send the third search result to the terminal device.

[0118] The following will refer to Figure 4 , taking intelligent question and answer search as an application scenario as an example, the entire information search process of the information search method of the exemplary implementation mode of the present disclosure is described in detail.

[0119] Reference Figure 4 As shown, in step 1, request acquisition, that is, the corresponding server receives an information search request, and the information search request at least includes the input information to be searched.

[0120] In the game scenario of intelligent question-answering search, the information to be searched is the question entered by the user.

[0121] Step 2: Search mode identification.

[0122] Exemplarily, the search mode is to determine the information category corresponding to the information to be searched. By determining the information category, it can be determined whether the information to be searched contained in the information search request needs to be searched by calling a search engine.

[0123] When it is determined based on the information category that it does not need to be searched, that is, the information category of the information to be searched is the third information category, step 3 is executed to directly obtain the question and answer result.

[0124] In step 3, the server responds that the information category corresponding to the information to be searched is the third information category, and inputs the information to be searched into the large language LLM model to obtain the third search result. In the process of executing this step, it is necessary to construct a prompt word (i.e., Prompt construction) for the information to be searched, and then perform LLM reasoning service on the information to be searched after the prompt word construction to obtain the final third search result.

[0125] On the contrary, when it is determined that the information needs to be searched based on the information category, step 4, question rewriting, is executed. That is, the information to be searched is preprocessed to obtain the preprocessed information to be searched; the information category is determined based on the preprocessed information to be searched, and then it is determined whether to conduct an external search.

[0126] The preprocessing method includes one or more of the following: converting the information format of the information to be searched; expanding the information content of the information to be searched.

[0127] When it is determined that no external search is required and only internal search is required, that is, in response to the information category corresponding to the information to be searched including the first information category, step 8 is executed to call the local search engine, that is, the first search result is obtained by calling the local search engine to search.

[0128] Exemplarily, when searching by calling a local search engine, the search can be performed by detecting similarity based on the vectorization processing, keyword extraction and other methods in the above embodiments, and the search results can be sorted.

[0129] Then continue to execute step 9, obtaining information content, that is, obtaining the content of the first search result and performing operations such as event extraction to provide basic data organization for generating the final result.

[0130] Finally, execute step 10 to obtain the question and answer results.

[0131] In this example, the first search result is constructed with prompt words, and the search result after the prompt word construction is input into the large language LLM model to generate the final second target search result.

[0132] When it is determined that an external search is required but an internal search is not required, only in response to the information category corresponding to the information to be searched containing the second information category, an external search engine is called to perform a search to obtain a second search result.

[0133] In the embodiment, the external search engine is divided into a commercial paid search engine and a metadata search engine. If a commercial paid search engine is required, step 5 is executed to call the commercial search engine; otherwise, step 6 is executed to call the metadata search engine.

[0134] The second search result obtained based on the above external search engine usually includes several web page titles, snapshots and URLs, etc. The obtained web pages are traversed, and the accessibility of the URL is first detected, and inaccessible web pages are discarded. Then, the web page content is obtained based on the HTML parsing plug-in. Next, the web page content is cleaned to remove impurities or unnecessary information. Finally, the web page summary and event content are extracted. This step provides basic data organization for generating the final result.

[0135] Finally, the above search results are also executed in step 10 to be input into the LLM model. The powerful semantic understanding and information analysis capabilities of the large language model are used to analyze the search results, extract content, reorganize, generate answers and citations, and make the returned answers more logical, organized and personalized, thereby effectively improving the user search experience.

[0136] When it is determined that both external and internal searches are required, the first search result obtained by calling the local search engine and the second search result obtained by calling the external search engine can be integrated to obtain the first target search result corresponding to the information to be searched, and the first target search result can be input into the LLM model to obtain the final second target search result.

[0137] In order to implement the above information search method, an information search device is provided in one embodiment of the present disclosure. Figure 5 A schematic architecture diagram of an information search device is shown.

[0138] The information search device 500 includes a request receiving module 501 , a category confirmation module 502 , a first search module 503 , a second search module 504 and a result determination module 505 .

[0139] The request receiving module 501 is used to receive an information search request, which at least includes input information to be searched; the category confirmation module 502 is used to determine the information category corresponding to the information to be searched; the first search module 503 is used to call a local search engine to search and obtain a first search result in response to the information category corresponding to the information to be searched including a first information category; the second search module 504 is used to call an external search engine to search and obtain a second search result in response to the information category corresponding to the information to be searched including a second information category; the result determination module 505 is used to determine a first target search result corresponding to the information to be searched based on the first search result and the second search result.

[0140] In an optional embodiment of the present disclosure, the category confirmation module 502 is used to preprocess the information to be searched to obtain the preprocessed information to be searched; determine the information category corresponding to the preprocessed information to be searched;

[0141] The preprocessing method includes one or more of the following: converting the information format of the information to be searched; expanding the information content of the information to be searched.

[0142] In an optional embodiment of the present disclosure, the external search engine at least includes a metadata search engine, and the metadata search engine includes multiple sub-search engines; the second search module 504 is used to call the metadata search engine in response to the information category corresponding to the information to be searched including the second information category, so as to send the information to be searched to multiple sub-search engines for searching based on the metadata search engine, and obtain search results corresponding to the multiple sub-search engines; the search results corresponding to the multiple sub-search engines are integrated to obtain a second search result.

[0143] In an optional embodiment of the present disclosure, the device may also include a processing module, the processing module is used to perform vectorization processing on the information to be searched in the first information category, or to perform keyword extraction on the information to be searched in the first information category, to obtain keywords corresponding to the information to be searched; a first search module 503 is used to perform a search based on the similarity between the information to be searched after vectorization processing and each search content in the local search engine, or based on the similarity between the keywords corresponding to the information to be searched and each search content in the local search engine, to obtain a first search result.

[0144] In an optional embodiment of the present disclosure, the device may further include an input module, which is used to construct a prompt word for the first target search result, and input the search result after the prompt word construction into the large language LLM model to generate a second target search result.

[0145] In an optional embodiment of the present disclosure, the processing module is also used to associate the information to be searched with the second target search results to generate first citation information; associate each search result in the second target search results with the corresponding search engine to generate second citation information; and generate citation information based on the first citation information and / or the second citation information so as to synchronously display the second search results and the corresponding citation information.

[0146] In an optional embodiment of the present disclosure, the input module is used for inputting the information to be searched into the large language model (LLM) to obtain a third search result in response to the information category corresponding to the information to be searched being the third information category.

[0147] The information search device 500 provided in the embodiment of the present disclosure can execute the technical solution of the information search method in any of the above-mentioned embodiments. Its implementation principle and beneficial effects are similar to the implementation principle and beneficial effects of the information search method. Please refer to the implementation principle and beneficial effects of the information search method, and no further details will be given here.

[0148] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the above method of the present specification is stored. In some possible implementations, various aspects of the present invention may also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps according to various exemplary embodiments of the present invention described in the above "Exemplary Method" section of the present specification.

[0149] The program product for implementing the above method according to the embodiment of the present invention may adopt a portable compact disc read-only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto, and in this document, a readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus, or device.

[0150] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0151] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0152] The program code contained in the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, radio frequency (RF), etc., or any suitable combination of the above.

[0153] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, and the like, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0154] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided. It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as a system, method or program product. Therefore, various aspects of the present invention may 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 may be collectively referred to herein as a "circuit", "module" or "system".

[0155] Refer to the following Figure 6 The electronic device 600 according to this embodiment of the present invention is described. Figure 6 The electronic device 600 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0156] like Figure 6As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different system components (including the storage unit 620 and the processing unit 610), and a display unit 640.

[0157] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 performs the steps according to various exemplary embodiments of the present invention described in the above “Exemplary Method” section of this specification. For example, the processing unit 610 can perform the following steps: Figure 2 Steps 201 to 205 shown in FIG.

[0158] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .

[0159] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0160] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0161] The electronic device 600 may also communicate with one or more external devices 1000 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 660. As shown, the network adapter 660 communicates with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0162] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0163] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0164] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0165] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

[0166] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An information search method, characterized in that: include: receiving an information search request, wherein the information search request at least includes input information to be searched; Determine the information category corresponding to the information to be searched; In response to the information category corresponding to the information to be searched including the first information category, calling a local search engine to search and obtain a first search result; In response to the information category corresponding to the information to be searched including the second information category, calling an external search engine to search and obtain a second search result; A first target search result corresponding to the information to be searched is determined according to the first search result and the second search result.

2. The method according to claim 1, characterized in that: Determining the information category corresponding to the information to be searched includes: Preprocessing the information to be searched to obtain preprocessed information to be searched; Determining the information category corresponding to the preprocessed information to be searched; Wherein, the pretreatment method comprises one or more of the following: Converting the information to be searched into an information format; The information to be searched is expanded in terms of information content.

3. The method according to claim 1, characterized in that The external search engine at least includes a metadata search engine, and the metadata search engine includes a plurality of sub-search engines; in response to the information category corresponding to the information to be searched including the second information category, calling the external search engine to search and obtain a second search result, including: In response to the information category corresponding to the information to be searched including the second information category, calling the metadata search engine to send the information to be searched to multiple sub-search engines for searching based on the metadata search engine, and obtaining search results corresponding to the multiple sub-search engines; The search results corresponding to the multiple sub-search engines are integrated to obtain the second search result.

4. The method according to claim 1, characterized in that: The method further comprises: Performing vectorization processing on the information to be searched in the first information category, or performing keyword extraction on the information to be searched in the first information category, to obtain keywords corresponding to the information to be searched; The local search engine is called to search and the first search result is obtained, including: The first search result is obtained by searching based on the similarity between the information to be searched after vectorization processing and each search content in the local search engine, or based on the similarity between the keywords corresponding to the information to be searched and each search content in the local search engine.

5. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: The first target search result is constructed with prompt words, and the search result after the prompt word construction is input into the large language LLM model to generate a second target search result.

6. The method according to claim 5, characterized in that The method further comprises: Associating the information to be searched with the second target search result to generate first citation information; Associating each search result in the second target search result with a corresponding search engine to generate second citation information; Based on the first citation information and / or the second citation information, citation information is generated so as to synchronously display the second search result and the corresponding citation information.

7. The method according to claim 1, characterized in that The method further comprises: In response to the information category corresponding to the information to be searched being the third information category, the information to be searched is input into the large language model (LLM) to obtain a third search result.

8. An information search device, characterized in that: The device comprises: A request receiving module, used for receiving an information search request, wherein the information search request at least includes input information to be searched; A category confirmation module, used to determine the information category corresponding to the information to be searched; A first search module, configured to call a local search engine to perform a search and obtain a first search result in response to the information category corresponding to the information to be searched including the first information category; A second search module, configured to call an external search engine to perform a search in response to the information category corresponding to the information to be searched including a second information category, and obtain a second search result; A result determination module is used to determine a first target search result corresponding to the information to be searched according to the first search result and the second search result.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the information search method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; Wherein, the processor is configured to perform the information search method according to any one of claims 1 to 7 by executing the executable instructions.

Images