Searching method and device, equipment and medium

Abstract

The invention provides a search method and device, equipment and a medium, and relates to the technical field of search, in particular to the technical field of artificial intelligence, intelligent agents and generative search. According to the implementation scheme, in response to a received target search request, a historical search request corresponding to the target search request is determined, and the relevancy between the historical search request and the target search request is not smaller than a preset threshold value; acquiring historical search data corresponding to the historical search request, the historical search data including interactive behavior information of a historical user for a historical search result of the historical search request; determining a target search intention based on the target search request and the historical search data; and determining a target search result based on the target search intention.

Description

Technical Field

[0001] The present disclosure relates to the field of search technology, in particular to the field of artificial intelligence, intelligent agents and generative search technology, and specifically to a search method, device, electronic device, computer-readable storage medium and computer program product. Background Art

[0002] Artificial intelligence is a discipline that studies how to use computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, planning, etc.). It includes both hardware-level and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing; artificial intelligence software technologies mainly include computer vision technology, speech recognition technology, natural language processing technology, as well as machine learning / deep learning, big data processing technology, knowledge graph technology, and other major directions.

[0003] With the development of computer technology, people rely on search engines to retrieve all kinds of information. In order to improve the user experience of using search engines, it is necessary to more accurately identify the user's search intent and return more comprehensive and accurate search results to the user.

[0004] The methods described in this section are not necessarily methods that have been previously conceived or employed. Unless otherwise indicated, it should not be assumed that any method described in this section is considered to be prior art simply because it is included in this section. Similarly, unless otherwise indicated, the issues mentioned in this section should not be considered to have been recognized in any prior art. Summary of the invention

[0005] The present disclosure provides a search method, apparatus, electronic device, computer-readable storage medium, and computer program product.

[0006] According to one aspect of the present disclosure, a search method is provided, comprising: in response to receiving a target search request, determining a historical search request corresponding to the target search request, wherein the correlation between the historical search request and the target search request is not less than a preset threshold; obtaining historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users for historical search results of the historical search request; determining a target search intent based on the target search request and the historical search data; and determining a target search result based on the target search intent.

[0007] According to another aspect of the present disclosure, a search device is provided, including: a first determination unit, configured to determine, in response to receiving a target search request, a historical search request corresponding to the target search request, wherein the correlation between the historical search request and the target search request is not less than a preset threshold; an acquisition unit, configured to acquire historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users for historical search results of the historical search request; a second determination unit, configured to determine a target search intent based on the target search request and the historical search data; and a third determination unit, configured to determine a target search result based on the target search intent.

[0008] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the above-mentioned search method.

[0009] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the above-mentioned search method.

[0010] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein the computer program can implement the above-mentioned search method when executed by a processor.

[0011] According to one or more embodiments of the present disclosure, the user's search intent may be more accurately identified, thereby improving search accuracy.

[0012] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.

[0014] Figure 1 A schematic diagram showing an exemplary system in which the various methods described herein may be implemented according to an exemplary embodiment of the present disclosure;

[0015] Figure 2 A flow chart of a search method according to an exemplary embodiment of the present disclosure is shown;

[0016] Figure 3 A schematic diagram showing an intention recognition process according to an exemplary embodiment of the present disclosure is shown;

[0017] Figure 4 A schematic diagram showing a structure of search results according to an exemplary embodiment of the present disclosure is shown;

[0018] Figure 5 A schematic diagram showing a search process according to an exemplary embodiment of the present disclosure is shown;

[0019] Figure 6 A structural block diagram of a search device according to an exemplary embodiment of the present disclosure is shown;

[0020] Figure 7 A structural block diagram of an exemplary electronic device that can be used to implement the embodiments of the present disclosure is shown. DETAILED DESCRIPTION

[0021] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0022] In the present disclosure, unless otherwise specified, the use of the terms "first", "second", etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of these elements, and such terms are only used to distinguish one element from another element. In some examples, the first element and the second element may refer to the same instance of the element, and in some cases, based on the description of the context, they may also refer to different instances.

[0023] The terms used in the description of various examples in this disclosure are only for the purpose of describing specific examples and are not intended to be limiting. Unless the context clearly indicates otherwise, if the number of elements is not specifically limited, the element can be one or more. In addition, the term "and / or" used in this disclosure covers any one of the listed items and all possible combinations.

[0024] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0025] Figure 1FIG. 1 is a schematic diagram of an exemplary system 100 in which various methods and apparatuses described herein may be implemented according to an embodiment of the present disclosure. Figure 1 , the system 100 includes one or more client devices 101, 102, 103, 104, 105, and 106, a server 120, and one or more communication networks 110 coupling the one or more client devices to the server 120. The client devices 101, 102, 103, 104, 105, and 106 may be configured to execute one or more applications.

[0026] In an embodiment of the present disclosure, the server 120 may run one or more services or software applications that enable the search method to be performed.

[0027] In some embodiments, server 120 may also provide other services or software applications, which may include non-virtualized environments and virtualized environments. In some embodiments, these services may be provided as web-based services or cloud services, such as provided to users of client devices 101, 102, 103, 104, 105, and / or 106 under a software as a service (SaaS) model.

[0028] exist Figure 1 In the configuration shown, the server 120 may include one or more components that implement the functions performed by the server 120. These components may include software components, hardware components, or a combination thereof that can be executed by one or more processors. Users operating client devices 101, 102, 103, 104, 105, and / or 106 may in turn utilize one or more client applications to interact with the server 120 to utilize the services provided by these components. It should be understood that a variety of different system configurations are possible, which may differ from the system 100. Therefore, Figure 1 is one example of a system for implementing the various methods described herein and is not intended to be limiting.

[0029] The user may use client devices 101, 102, 103, 104, 105 and / or 106 to send a search request. The client device may provide an interface that enables the user of the client device to interact with the client device. The client device may also output information to the user via the interface. Figure 1 Only six client devices are depicted, but one skilled in the art will appreciate that the present disclosure may support any number of client devices.

[0030] Client devices 101, 102, 103, 104, 105 and / or 106 may include various categories of computer devices, such as portable handheld devices, general-purpose computers (such as personal computers and laptop computers), workstation computers, wearable devices, smart screen devices, self-service terminal devices, service robots, game systems, thin clients, various messaging devices, sensors or other sensing devices, etc. These computer devices may run various categories and versions of software applications and operating systems, such as MICROSOFT Windows, APPLE iOS, UNIX-like operating systems, Linux or Linux-like operating systems (such as GOOGLE Chrome OS); or include various mobile operating systems, such as MICROSOFT Windows Mobile OS, iOS, Windows Phone, Android. Portable handheld devices may include cellular phones, smart phones, tablet computers, personal digital assistants (PDAs), etc. Wearable devices may include head-mounted displays (such as smart glasses) and other devices. Game systems may include various handheld game devices, Internet-enabled game devices, etc. The client device is capable of executing a variety of different applications, such as various Internet-related applications, communication applications (eg, email applications), short message service (SMS) applications, and may use various communication protocols.

[0031] The network 110 may be any type of network known to those skilled in the art that may support data communications using any of a variety of available protocols, including but not limited to TCP / IP, SNA, IPX, etc. By way of example only, one or more networks 110 may be a local area network (LAN), an Ethernet-based network, a token ring, a wide area network (WAN), the Internet, a virtual network, a virtual private network (VPN), an intranet, an extranet, a blockchain network, a public switched telephone network (PSTN), an infrared network, a wireless network (e.g., Bluetooth, WIFI), and / or any combination of these and / or other networks.

[0032] Server 120 may include one or more general purpose computers, dedicated server computers (e.g., PC (personal computer) servers, UNIX servers, mid-range servers), blade servers, mainframe computers, server clusters, or any other suitable arrangement and / or combination. Server 120 may include one or more virtual machines running virtual operating systems, or other computing architectures involving virtualization (e.g., one or more flexible pools of logical storage devices that may be virtualized to maintain a server's virtual storage device). In various embodiments, server 120 may run one or more services or software applications that provide the functionality described below.

[0033] The computing units in the server 120 may run one or more operating systems including any of the above operating systems and any commercially available server operating systems. The server 120 may also run any of a variety of additional server applications and / or middle-tier applications, including HTTP servers, FTP servers, CGI servers, JAVA servers, database servers, etc.

[0034] In some implementations, server 120 may include one or more applications to analyze and consolidate data feeds and / or event updates received from users of client devices 101, 102, 103, 104, 105, and 106. Server 120 may also include one or more applications to display data feeds and / or real-time events via one or more display devices of client devices 101, 102, 103, 104, 105, and 106.

[0035] In some embodiments, the server 120 may be a server of a distributed system, or a server combined with a blockchain. The server 120 may also be a cloud server, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology. A cloud server is a host product in a cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and virtual private servers (VPS) services.

[0036] The system 100 may also include one or more databases 130. In some embodiments, these databases may be used to store data and other information. For example, one or more of the databases 130 may be used to store information such as audio files and video files. The databases 130 may reside in various locations. For example, the database used by the server 120 may be local to the server 120, or may be remote from the server 120 and may communicate with the server 120 via a network-based or dedicated connection. The databases 130 may be of different categories. In some embodiments, the databases used by the server 120 may be, for example, relational databases. One or more of these databases may store, update, and retrieve data to and from the databases in response to commands.

[0037] In some embodiments, one or more of the databases 130 may also be used by applications to store application data. The databases used by the applications may be different categories of databases, such as a key-value store, an object store, or a conventional store backed by a file system.

[0038] Figure 1The system 100 may be configured and operated in various ways to enable the application of various methods and apparatuses described in the present disclosure.

[0039] In the related art, semantic recognition is usually performed based on the search request input by the user, and then the search results are recalled based on the relevance between the database content and the search request, which fails to accurately identify the user's search intention.

[0040] The inventors have found through research that after a user inputs a search request and browses the search results, by analyzing the user's behavior data on the search results, it is possible to determine whether the search results displayed to the user can meet the user's search intent. For example, when a historical user clicks on link B in the search results page after searching based on search request A and browses for a long time, it indicates that the content of link B can meet the search request of the historical user, and the corresponding search intent of search request A can be determined based on the content of link B.

[0041] Based on this, the present disclosure provides a search method which, after receiving a target search request, performs intent recognition based on historical search data associated with the target search request, thereby being able to use the historical search data to more accurately determine the target search intent, and then determine target search results that are more in line with the user's intent, thereby improving search accuracy.

[0042] Figure 2 FIG. 2 shows a flow chart of a search method 200 according to an exemplary embodiment of the present disclosure. Figure 2 As shown, the method 200 includes:

[0043] Step S201: in response to receiving a target search request, determining a historical search request corresponding to the target search request, wherein a correlation between the historical search request and the target search request is not less than a preset threshold;

[0044] Step S202: Acquire historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users with respect to historical search results of the historical search request;

[0045] Step S203: determining a target search intention based on the target search request and the historical search data; and

[0046] Step S204: Determine target search results based on the target search intent.

[0047] By applying the above-mentioned search method 200, after receiving a target search request (i.e., the query information input by the user), intent recognition can be performed based on the historical search data associated with the target search request, and the historical search data can be used to more accurately determine the target search intent, thereby determining a target search result that is more in line with the user's intent, thereby improving search accuracy.

[0048] In some examples, the historical search data may be obtained based on the behavior data of multiple historical users in the historical search process, and the multiple historical users may include the user who is currently searching. In some examples, the historical search database may be constructed after the historical search behavior data of multiple historical users are counted and aggregated. In this case, the historical search database may include multiple historical search requests and the historical search data corresponding to each historical search request.

[0049] In some examples, step S201 may be based on the currently received target search request to search in the above-mentioned historical search database to obtain historical search requests related to the target search request. In some examples, the relevance between the target search request and the historical search request may be determined based on various methods, such as determining the semantic feature vector corresponding to each search request and calculating the cosine similarity between the semantic feature vectors, or calculating the similarity based on the word vector corresponding to each search request. As long as the historical search request related to the target search request can be retrieved, the present disclosure does not limit the specific retrieval method.

[0050] According to some embodiments, the historical search data includes at least one of the following: first information of the historical user's click behavior or browsing behavior for the historical search results; second information of the historical user adjusting the historical search request after browsing the historical search results; and third information of the historical user's click behavior or browsing behavior for the adjusted search results of the historical search request. Thus, the user's search intent can be indicated based on the user's click or browsing behavior for the historical search results corresponding to the historical search request, or the behavior of adjusting the historical search request, or the click or browsing behavior after adjusting the historical search request, so as to achieve more accurate intent recognition and thus improve search accuracy.

[0051] As described above, by analyzing the user's behavior data on search results, it can be determined whether the search results displayed to the user can meet the user's search intent. For example, when a historical user only clicks on a link in the search results page and browses for a long time after searching based on search request A, it can be determined that the search results page returned by the search engine based on search request A can meet the user's search intent. In some examples, the search results page also includes summary information. When the user ends the search after only browsing the search results page, it can be determined that the summary information can meet the user's search intent.

[0052] In some examples, when a user adjusts a historical search request after browsing historical search results, the user's search intent can be determined based on the user's adjustment behavior information. For example, when a historical user adjusts the search request content to ABD after searching based on search request ABC and browsing the search results, it can be determined that ABD is more indicative of the user's search intent than ABC. Furthermore, similar to the principle described above of determining whether the initial search results can satisfy the user's search intent based on the user's click or browsing behavior information for the initial search results, it is also possible to determine whether the adjusted search results can satisfy the user's search intent based on the user's click or browsing behavior information for the adjusted search results.

[0053] In some examples, the historical search data of each historical search request may include one or more of the following: search volume, click-through rate of the search results page, dwell time on the search results page, click-through rate for each link in the search results page, title of one or more links with the highest click-through rate in the search results page, number of times the historical user adjusted the historical search request after browsing the search results page, information on the adjusted content of the historical search request by the historical user, number of times the historical user adjusted the historical search request, information on the click or browsing behavior of the historical user for the adjusted search results (for example, it may include click-through rate, dwell time, title of one or more links with the highest click-through rate, etc.), information on other historical search requests associated with the historical search request (for example, when the historical search request is a certain entity word, other historical search requests associated with it may be historical search requests including the entity word). In some examples, the search results page may include multiple types of result cards, for example, an Aladdin search card including an intent tag, in which case the historical search data may include the explicit intent tag corresponding to the Aladdin search card clicked or browsed by the user.

[0054] According to some embodiments, determining the target search intent based on the target search request and the historical search data in step S203 includes: generating first input data based on the target search request and the historical search data; inputting the first input data into an intent recognition model to obtain the target search intent output by the intent recognition model, wherein the intent recognition model is trained using first sample data including a sample search request and sample search data. Thus, the target search intent can be determined using the intent recognition model, and by using the trained intent recognition model to determine the target search intent corresponding to the target search data, the efficiency and accuracy of intent recognition can be improved.

[0055] In some examples, the intent recognition model can be trained in the following manner: first sample data including a sample search request and sample search data is input into the intent recognition model to obtain a predicted search intent, a predicted score is determined based on the predicted search intent, a loss value is calculated based on the predicted score, and the parameters of the intent recognition model are adjusted based on the loss value.

[0056] In some examples, the first sample data may further include a reference search intent. In this case, after the predicted search intent is obtained using the intent recognition model, a loss value can be calculated based on the reference search intent and the predicted search intent, and then the parameters of the intent recognition model can be adjusted based on the loss value.

[0057] In some examples, sample search requests and sample search data corresponding to the sample search requests can be determined based on historical search data of real users, and the sample search requests and sample search data can be input into a generative language model to determine a reference search intent, thereby constructing a first sample data including the sample search request, sample search data, and reference search intent. The natural language processing capabilities of the generative language model can be used to more conveniently and efficiently construct training data for the intent recognition model, thereby improving the training efficiency of the intent recognition model.

[0058] According to some embodiments, the intent recognition model is obtained by fine-tuning the first generative language model using the first sample data. By fine-tuning the pre-trained first generative language model to obtain the intent recognition model, the natural language processing capabilities of the pre-trained language model can be effectively utilized to improve the training efficiency and prediction performance of the intent recognition model.

[0059] According to some embodiments, the historical search data includes first information of the historical user's click behavior on the historical search results and second information of the historical user adjusting the historical search request after browsing the historical search results, and the generation of the first input data based on the target search request and the historical search data includes: determining the target information from the first information and the second information based on the data volume of the first information and the second information; and generating the first input data based on the target search request and the target information. By selecting more important target information based on the data volume, it is possible to filter the historical search data, reduce the data volume of the first input data while ensuring the importance of the data, save hardware resources, and improve search efficiency.

[0060] It can be understood that the larger the amount of historical search data corresponding to the historical search request, the higher the confidence of the historical search data. In some examples, when the click behavior information corresponding to the historical search request is relatively rich (large amount of data), the first information can be selected as the target information to accurately represent the search intention of the historical user. In some examples, when the information of adjusting the historical search request corresponding to the historical search request is relatively rich, the second information can be selected as the target information to accurately represent the search intention of the historical user, thereby achieving more accurate intent recognition.

[0061] According to some embodiments, determining the target search result based on the target search intent in step S204 includes: determining the target search result based on the target search intent using a second generative language model. By using a generative large language model to obtain the target search result, the generation quality of the search result can be improved, and the user's search intent can be more accurately met.

[0062] In some examples, content generation instructions may be determined based on the target search intent so that the second generative model can generate content results that meet the target search intent based on the content generation instructions, thereby improving the efficiency of generating target search results.

[0063] In some examples, similar to the training process of the intent recognition model described above, the second generative language model can also be fine-tuned based on the pre-trained language model, for example, the supervised fine-tuning (SFT) method can be applied to train the pre-trained large model. The sample corpus used to train the second generative language model may include sample input data and reference output data. During the training process, the sample input data may be input into the second generative language model to obtain predicted output data, the loss value may be calculated based on the predicted output data and the sample output data, and then the model may be fine-tuned in all parameters based on the loss value.

[0064] In some examples, the first generative language model and the second generative language model may be the same. In some examples, the two may be different. For example, a smaller-scale language model may be selected as the first generative language model, and a larger-scale language model may be selected as the second generative language model, so that the occupation of hardware resources by the language model can be adapted to its performance requirements, thereby improving the efficiency of intent recognition and result generation.

[0065] In some examples, method 200 can be applied to the field of intelligent agents, that is, during the interaction between a user and an intelligent agent, a target search request is determined based on information sent by the user to the intelligent agent, and then by applying the above method 200, a generative language model corresponding to the intelligent agent is used to generate target search results and return them to the user.

[0066] According to some embodiments, the determining the target search result based on the target search intent using the second generative language model includes: searching based on the target search intent to obtain a first search result; inputting second input data including the target search intent and the first search result into the second generative language model; and determining the target search result based on output data of the second generative language model. Thus, richer and more comprehensive output data can be obtained by embedding preliminary search results in the input data, so that the target search result better meets the needs of users and improves the user experience.

[0067] In some examples, the above technical means can be implemented through retrieval-augmented generation (RAG), that is, obtaining relevant information from an external knowledge base and embedding it into the input data for inputting a generative language model to obtain more comprehensive output data. In this example, the external knowledge base can be pre-screened and built, for example, a document vector index can be constructed for the screened high-quality documents. After keyword extraction based on the target search intent, document vectors can be recalled in the knowledge base through ANN (Approximate Nearest Neighbor) retrieval, and then filtered based on the relevance of the recall results (preliminary search results) to the target search intent, and the second input data is determined using the screened first search result.

[0068] According to some embodiments, the second input data is determined by embedding the target search intent and the first search result into an input data template to obtain the second input data, wherein the input data template includes task description information, and the task description information indicates that the second input data includes the target search intent and the first search result. By generating the second input data based on the formatted input data template, the content form of the generated result can be more accurately controlled, and the generation quality of the target search result can be improved.

[0069] According to some embodiments, the historical search results are obtained by inputting the historical search request into a search engine, and the method 200 further includes: determining a diagnosis result for the historical search results based on the target search intent and the historical search data, wherein the diagnosis result indicates whether the historical search results can meet the target search intent, wherein the step S204 determines the target search results based on the target search intent, including: in response to determining that the diagnosis result indicates that the historical search results cannot meet the target search intent, using the second generative language model to determine the target search results based on the target search intent; and in response to determining that the diagnosis result indicates that the historical search results cannot meet the target search intent, inputting the target search request into the search engine to obtain the target search results. Thus, it is possible to determine whether the current search capability of the search engine (i.e., the result currently output by the search engine) can meet the needs of the user based on the historical search data, and when the diagnosis result indicates that the current search capability of the search engine cannot meet the search intent of the user, the generation capability of the generative language model is applied to improve the quality of the search results, save hardware resources, and improve search efficiency.

[0070] In some examples, the diagnosis result can also be obtained using a generative language model. For example, after fine-tuning the first generative language model to obtain a language model with intent recognition capability, the target search request and historical search data are input into the fine-tuned first generative language model in step S203, so that the first generative language model can output the target search intent and simultaneously output the diagnosis result determined based on the target search intent. In this example, the sample data used to fine-tune the first generative language model can also include annotation information of the reference diagnosis result, so that the fine-tuned first generative language model can simultaneously perform the intent recognition task and the diagnosis task of diagnosing whether the search results can meet the target search intent, so as to improve the search efficiency.

[0071] According to some embodiments, the target search intent includes multiple sub-intents, and determining the target search result based on the target search intent in step S204 includes: searching based on each of the multiple sub-intents to obtain multiple second search results; and determining the target search result based on the multiple second search results. By performing multi-target search based on multiple sub-intents, more comprehensive search results can be obtained, which can meet the user's search needs as much as possible and improve the user experience.

[0072] Figure 3 FIG. 2 is a schematic diagram showing an intention recognition process according to an exemplary embodiment of the present disclosure. Figure 3 As described above, after receiving the target search request input by the user, the historical search data corresponding to the historical search request related to the target search request can be retrieved and determined in the search behavior database 301 based on the target search request. In this example, the search behavior database 301 can be constructed by counting and aggregating the historical search behavior data of multiple historical users based on the method described above. In this case, the search behavior database 301 can include multiple historical search requests and the historical search data corresponding to each historical search request.

[0073] After determining the historical search data corresponding to the target search request, the first input data can be generated based on the target search request and the historical search data and input into the intention recognition model 302 to obtain the target search intent output by the intention recognition model 302 and the diagnostic results indicating whether the search results can meet the target search intent, that is, the intention recognition model 302 is used to achieve simple and fast intent recognition and search result diagnosis, thereby improving search efficiency.

[0074] Figure 4 A schematic diagram of the structure of search results according to an exemplary embodiment of the present disclosure is shown. Figure 5 Schematic diagram of a search process according to an exemplary embodiment of the present disclosure is shown. In this example, the target search result may be a search result page 400 obtained using a generative language model. Figure 4 As shown, in this example, search result page 400 includes summary content area 401 , link function area 402 , intent directory area 403 , main content area 404 , and reference content area 405 .

[0075] In this example, the summary information contained in the summary content area 401 can be obtained based on the summary generation instruction using the generative language model 503. As described above, the target search intent corresponding to the target search request can be obtained by executing steps S201-S203 described above. For example, based on the target search request, historical search data corresponding to historical search requests related to the target search request can be retrieved and determined in the search behavior database 501, and then the target search request and the historical search data are input into the intent recognition model 502 to obtain the target search intent output by the intent recognition model 302. In this example, the target search intent includes N sub-intentions. In this case, M retrieval results can be determined based on multiple sub-intentions. It can be understood that each sub-intention can correspond to one or more retrieval results, for example, it can correspond to multiple different types of retrieval results (documents, videos, etc.), and M and N can be equal or different.

[0076] Continue to see Figure 4 and Figure 5 , when multiple sub-intents and multiple search results are determined, the link function area 402 can be used to display the index link of the key result, so that the user can directly click the index link in the search result page to view the key result, and the reference content area 405 can be used to display the reference content, so that the user can view the reference basis of each content in the search result page 400. The display content of the intent directory area 403 can be determined based on multiple sub-intents, and the content structure of the search result page can be organized more clearly by displaying the intent directory to the user. For example, multiple content generation instructions can be determined based on multiple sub-intents, so that multiple groups of content results corresponding to multiple sub-intents can be obtained based on multiple content generation instructions using the generative language model 503. On this basis, the generated content results can be integrated based on the content integration instruction using the generative language model 503, and then multiple groups of content results and the corresponding integrated results of the multiple groups of content results can be displayed in the main content area 404.

[0077] In some examples, the multiple groups of content results corresponding to multiple sub-intents can be multimodal information, for example, they can include web pages, documents, videos, pictures, and other forms, and can also include various service links, product links, and other consumer content to fully meet user needs and improve the conversion rate of online services or e-commerce products.

[0078] By applying the above-mentioned search process, after receiving a target search request (i.e., the query information input by the user), intent recognition can be performed based on the historical search data associated with the target search request, and the historical search data can be used to more accurately determine the target search intent, thereby determining target search results that are more in line with the user's intent, thereby improving search accuracy.

[0079] According to one aspect of the present disclosure, a search device is also provided. Figure 6 FIG. 6 shows a structural block diagram of a search device 600 according to an exemplary embodiment of the present disclosure. Figure 6 As shown, the device 600 includes:

[0080] A first determining unit 601 is configured to, in response to receiving a target search request, determine a history search request corresponding to the target search request, wherein the correlation between the history search request and the target search request is not less than a preset threshold;

[0081] An acquisition unit 602 is configured to acquire historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users with respect to historical search results of the historical search request;

[0082] A second determining unit 603 is configured to determine a target search intention based on the target search request and the historical search data; and

[0083] The third determining unit 604 is configured to determine a target search result based on the target search intention.

[0084] According to some embodiments, the historical search data includes at least one of the following: first information on the historical user's click behavior or browsing behavior with respect to the historical search results; second information on the historical user adjusting the historical search request after browsing the historical search results; and third information on the historical user's click behavior or browsing behavior with respect to the adjusted search results of the historical search request.

[0085] According to some embodiments, the second determination unit 603 includes: a generation subunit, configured to generate first input data based on the target search request and the historical search data; a first input subunit, configured to input the first input data into an intent recognition model to obtain the target search intent output by the intent recognition model, wherein the intent recognition model is trained using first sample data including a sample search request and sample search data.

[0086] According to some embodiments, the intent recognition model is obtained by fine-tuning and training a first generative language model using the first sample data.

[0087] According to some embodiments, the historical search data includes first information on the historical user's click behavior on the historical search results and second information on the historical user adjusting the historical search request after browsing the historical search results, and the generation subunit includes: a determination module, configured to determine target information from the first information and the second information based on the data volume of the first information and the second information; and a generation module, configured to generate the first input data based on the target search request and the target information.

[0088] According to some embodiments, the third determination unit 604 is configured to: determine the target search result based on the target search intent using a second generative language model.

[0089] According to some embodiments, the third determination unit 604 includes: a first search subunit, configured to search based on the target search intent to obtain a first search result; a second input subunit, configured to input second input data including the target search intent and the first search result into the second generative language model; and a first determination subunit, configured to determine the target search result based on output data of the second generative language model.

[0090] According to some embodiments, the second input data is determined by embedding the target search intent and the first search result into an input data template to obtain the second input data, wherein the input data template includes task description information, and the task description information indicates that the second input data includes the target search intent and the first search result.

[0091] According to some embodiments, the historical search results are obtained by inputting the historical search request into a search engine, and the device 600 also includes: a fourth determination unit, configured to determine a diagnosis result for the historical search results based on the target search intent and the historical search data, wherein the diagnosis result indicates whether the historical search results can meet the target search intent, wherein the third determination unit 604 is configured to: in response to determining that the diagnosis result indicates that the historical search results cannot meet the target search intent, determine the target search results based on the target search intent using the second generative language model; and in response to determining that the diagnosis result indicates that the historical search results cannot meet the target search intent, input the target search request into the search engine to obtain the target search results.

[0092] According to some embodiments, the target search intent includes multiple sub-intentions, and the third determination unit 604 includes: a second search sub-unit, configured to search based on each of the multiple sub-intentions to obtain multiple second search results; and a second determination sub-unit, configured to determine the target search result based on the multiple second search results.

[0093] It should be understood that Figure 6 The operations of the various units of the search device 600 shown in FIG. 6 can be similar to those of FIG. Figure 2 The steps in the search method 200 described above correspond to each other. Therefore, the operations, features and advantages described above for the method 200 are also applicable to the apparatus 600 and the various units included therein. For the sake of brevity, some operations, features and advantages are not described in detail here.

[0094] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0095] According to another aspect of the present disclosure, an electronic device is also provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the above-mentioned search method.

[0096] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is also provided, wherein the computer instructions are used to enable the computer to execute the above-mentioned search method.

[0097] According to another aspect of the present disclosure, a computer program product is also provided, including a computer program, wherein the computer program implements the above-mentioned search method when executed by a processor.

[0098] refer to Figure 7 , a block diagram of an electronic device 700 that can be used as a server or client of the present disclosure will now be described, which is an example of a hardware device that can be applied to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0099] like Figure 7 As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0100] Multiple components in the device 700 are connected to the I / O interface 705, including: an input unit 706, an output unit 707, a storage unit 708, and a communication unit 709. The input unit 706 can be any type of device that can input information to the device 700. The input unit 706 can receive input digital or character information and generate key signal input related to user settings and / or function control of the electronic device, and can include but is not limited to a mouse, a keyboard, a touch screen, a track pad, a track ball, a joystick, a microphone, and / or a remote controller. The output unit 707 can be any type of device that can present information, and can include but is not limited to a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 708 can include but is not limited to a disk, an optical disk. The communication unit 709 allows the device 700 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and can include but is not limited to a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a Bluetooth™ device, an 802.11 device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0101] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 701 performs the various methods and processes described above, such as search methods. For example, in some embodiments, the search method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via ROM 702 and / or communication unit 709. When the computer program is loaded into RAM 703 and executed by the computing unit 701, one or more steps of the search method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to perform the search method in any other appropriate manner (e.g., by means of firmware).

[0102] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0103] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0104] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer 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 foregoing.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0106] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0107] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0108] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0109] Although the embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it should be understood that the above-mentioned methods, systems and devices are merely exemplary embodiments or examples, and the scope of the present invention is not limited by these embodiments or examples, but only by the claims after authorization and their equivalent scope. Various elements in the embodiments or examples can be omitted or replaced by their equivalent elements. In addition, each step can be performed in an order different from that described in the present disclosure. Further, the various elements in the embodiments or examples can be combined in various ways. It is important that with the evolution of technology, many elements described herein can be replaced by equivalent elements that appear after the present disclosure.

Claims

1. A search method, comprising: In response to receiving a target search request, determining a historical search request corresponding to the target search request, wherein a correlation between the historical search request and the target search request is not less than a preset threshold; Acquire historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users with respect to historical search results of the historical search request; Determining a target search intent based on the target search request and the historical search data; and Based on the target search intent, target search results are determined.

2. The method of claim 1, wherein: The historical search data includes at least one of the following: first information of the historical user's click behavior or browsing behavior on the historical search results; The historical user adjusts the second information of the historical search request after browsing the historical search results; as well as The third information of the historical user's click behavior or browsing behavior on the adjusted search results of the historical search request.

3. The method according to claim 1 or 2, wherein: The determining the target search intention based on the target search request and the historical search data includes: generating first input data based on the target search request and the historical search data; The first input data is input into an intent recognition model to obtain the target search intent output by the intent recognition model, wherein the intent recognition model is trained using first sample data including a sample search request and sample search data.

4. The method of claim 3, wherein: The intent recognition model is obtained by fine-tuning and training the first generative language model using the first sample data.

5. The method according to claim 3 or 4, wherein: The historical search data includes first information of the historical user's click behavior on the historical search results and second information of the historical user adjusting the historical search request after browsing the historical search results. The generating of the first input data based on the target search request and the historical search data includes: determining target information from the first information and the second information based on data amounts of the first information and the second information; and The first input data is generated based on the target search request and the target information.

6. The method according to any one of claims 1 to 5, wherein: Determining the target search result based on the target search intent includes: The target search result is determined based on the target search intent using a second generative language model.

7. The method of claim 6, wherein: Determining the target search result based on the target search intent using the second generative language model includes: Searching based on the target search intent to obtain a first search result; inputting second input data including the target search intent and the first search result into the second generative language model; and The target search result is determined based on the output data of the second generative language model.

8. The method of claim 7, wherein: The second input data is determined in the following manner: The target search intent and the first search result are embedded in an input data template to obtain the second input data, wherein the input data template includes task description information, and the task description information indicates that the second input data includes the target search intent and the first search result.

9. The method according to any one of claims 6 to 8, wherein: The historical search results are obtained by inputting the historical search request into a search engine, and the method further comprises: Based on the target search intent and the historical search data, determining a diagnosis result for the historical search results, wherein the diagnosis result indicates whether the historical search results can satisfy the target search intent, Wherein, determining the target search result based on the target search intention includes: In response to determining that the diagnosis result indicates that the historical search results cannot satisfy the target search intent, determining the target search results based on the target search intent using the second generative language model; and In response to determining that the diagnosis result indicates that the historical search results cannot satisfy the target search intention, the target search request is input into the search engine to obtain the target search results.

10. The method according to any one of claims 1 to 8, wherein: The target search intent includes multiple sub-intents, and determining the target search result based on the target search intent includes: Searching based on each of the multiple sub-intentions to obtain multiple second search results; and Based on the plurality of second search results, the target search result is determined.

11. A search device, comprising: A first determining unit is configured to, in response to receiving a target search request, determine a history search request corresponding to the target search request, wherein a correlation between the history search request and the target search request is not less than a preset threshold; an acquisition unit, configured to acquire historical search data corresponding to the historical search request, wherein the historical search data includes interaction behavior information of historical users with respect to historical search results of the historical search request; A second determining unit is configured to determine a target search intention based on the target search request and the historical search data; and The third determining unit is configured to determine a target search result based on the target search intention.

12. The device of claim 11, wherein: The historical search data includes at least one of the following: first information of the historical user's click behavior or browsing behavior on the historical search results; The historical user adjusts the second information of the historical search request after browsing the historical search results; as well as The third information of the historical user's click behavior or browsing behavior on the adjusted search results of the historical search request.

13. The device according to claim 11 or 12, wherein: The second determining unit includes: A generating subunit, configured to generate first input data based on the target search request and the historical search data; The first input subunit is configured to input the first input data into an intent recognition model to obtain the target search intent output by the intent recognition model, wherein the intent recognition model is trained using first sample data including a sample search request and sample search data.

14. The device of claim 13, wherein: The intent recognition model is obtained by fine-tuning and training the first generative language model using the first sample data.

15. The device according to claim 13 or 14, wherein: The historical search data includes first information of the historical user's click behavior on the historical search results and second information of the historical user adjusting the historical search request after browsing the historical search results. The generating subunit includes: a determination module configured to determine target information from the first information and the second information based on data amounts of the first information and the second information; and A generating module is configured to generate the first input data based on the target search request and the target information.

16. The device according to any one of claims 11 to 15, wherein: The third determining unit is configured to: The target search result is determined based on the target search intent using a second generative language model.

17. The device of claim 16, wherein: The third determining unit includes: A first search subunit is configured to search based on the target search intent to obtain a first search result; A second input subunit is configured to input second input data including the target search intent and the first search result into the second generative language model; and The first determining subunit is configured to determine the target search result based on the output data of the second generative language model.

18. The device of claim 17, wherein: The second input data is determined in the following manner: The target search intent and the first search result are embedded in an input data template to obtain the second input data, wherein the input data template includes task description information, and the task description information indicates that the second input data includes the target search intent and the first search result.

19. The device according to any one of claims 16 to 18, wherein: The historical search results are obtained by inputting the historical search request into a search engine, and the device further comprises: a fourth determining unit, configured to determine a diagnosis result for the historical search results based on the target search intent and the historical search data, wherein the diagnosis result indicates whether the historical search results can satisfy the target search intent, Wherein, the third determining unit is configured as follows: In response to determining that the diagnosis result indicates that the historical search results cannot satisfy the target search intent, determining the target search results based on the target search intent using the second generative language model; and In response to determining that the diagnosis result indicates that the historical search results cannot satisfy the target search intention, the target search request is input into the search engine to obtain the target search results.

20. The device according to any one of claims 11 to 18, wherein: The target search intent includes multiple sub-intents, and the third determination unit includes: A second search subunit is configured to search based on each of the multiple sub-intentions to obtain multiple second search results; and The second determining subunit is configured to determine the target search result based on the multiple second search results.

21. An electronic device comprising: at least one processor; as well as a memory communicatively coupled to the at least one processor; in The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

22. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to make a computer execute the method according to any one of claims 1-10.

23. A computer program product comprising a computer program, wherein: The computer program implements the method according to any one of claims 1 to 10 when executed by a processor.

Images