Search Information Processing Method, Apparatus, Electronic Device, and Storage Medium

Abstract

The present disclosure provides a search information processing method, apparatus, electronic device, and storage medium, which relate to the field of artificial intelligence technology, and particularly to technical fields such as intelligent search, big data, deep learning, large language models, etc., and can be used in scenarios such as information retrieval and human-computer interaction. The specific implementation solution is as follows: in response to obtaining search information, use a large language model to process the search information to obtain multiple candidate search texts, where the candidate search texts include generated search information associated with the search information; perform semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts, where the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes; determine at least one target search text from the multiple candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts; display the target search text.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technologies, particularly to technologies such as intelligent search, big data, deep learning, large language models, etc., and can be used in scenarios such as information retrieval and human-computer interaction. Background Art

[0002] With the rapid development of Internet technologies, users can quickly browse resource information such as news through terminal devices such as smart phones, and can also input search information such as text in the terminal device based on needs to retrieve resource information. The terminal device can perform page jumps based on the search information to provide a page that matches the user's retrieval requirements. Summary of the Invention

[0003] The present disclosure provides a search information processing method, apparatus, electronic device, and storage medium.

[0004] According to one aspect of the present disclosure, there is provided a search information processing method, including: in response to obtaining search information, processing the search information using a large language model to obtain a plurality of candidate search texts, where the candidate search texts include generated search texts associated with the search information; performing semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts, where the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes; determining at least one target search text from the plurality of candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the plurality of candidate search texts; and displaying the target search text.

[0005] According to another aspect of the present disclosure, there is provided a search information processing apparatus, including: a candidate search text obtaining module, configured to process search information using a large language model in response to obtaining the search information to obtain a plurality of candidate search texts, where the candidate search texts include generated search texts associated with the search information; an associated text obtaining module, configured to perform semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts, where the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes; a target search text determining module, configured to determine at least one target search text from the plurality of candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the plurality of candidate search texts; and a display module, configured to display the target search text.

[0006] According to another aspect of the present disclosure, there is provided an electronic device, including: 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 to enable the at least one processor to execute the method provided according to an embodiment of the present disclosure.

[0007] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the method provided according to an embodiment of the present disclosure.

[0008] According to another aspect of the present disclosure, there is provided a computer program product, including a computer program, which implements the method provided according to an embodiment of the present disclosure when executed by a processor.

[0009] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understandable through the following description. Description of the Drawings

[0010] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0011] Figure 1 Schematically shows an exemplary system architecture to which the search information processing method and apparatus according to an embodiment of the present disclosure can be applied;

[0012] Figure 2 Schematically shows a flowchart of the search information processing method according to an embodiment of the present disclosure;

[0013] Figure 3 Schematically shows an application scenario diagram of the search information processing method according to an embodiment of the present disclosure;

[0014] Figure 4 Schematically shows an application scenario diagram of the search information processing method according to another embodiment of the present disclosure;

[0015] Figure 5 Schematically shows a block diagram of the search information processing apparatus according to an embodiment of the present disclosure; and

[0016] Figure 6 Schematically shows a block diagram of an electronic device suitable for implementing the search information processing method according to an embodiment of the present disclosure. Detailed Embodiments

[0017] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0018] In the technical solutions of the present disclosure, the acquisition, storage, and application of the user's personal information involved all comply with the provisions of relevant laws and regulations, necessary confidentiality measures are taken, and public order and good customs are not violated.

[0019] Embodiments of the present disclosure provide a search information processing method, apparatus, electronic device, and storage medium. The search information processing method includes: in response to obtaining search information, processing the search information using a large language model to obtain multiple candidate search texts, where the candidate search texts include generated search information associated with the search information; performing semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts, where the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes; determining at least one target search text from the multiple candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts; and displaying the target search text.

[0020] According to the embodiments of the present disclosure, by using a large language model to process search information and generating candidate search texts including the search information and the generated search information, the generated candidate search texts can supplement the obtained search information, and the multiple candidate search texts can more completely and accurately represent the actual retrieval requirements of the user. Performing semantic search based on the candidate search texts can retrieve associated texts that are semantically similar or identical to the candidate search texts from the preset search text library. By obtaining the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts through the retrieval requirement attribute contribution information of the associated texts, the contribution degree of the candidate search texts to the retrieval requirement attributes can be evaluated, and then the target search text can be determined according to the retrieval requirement attribute contribution information, so that the target search text can more accurately meet the retrieval requirements of the target object, and the target object can quickly and conveniently obtain the information to be retrieved based on the displayed target search text, improving the retrieval efficiency and retrieval accuracy.

[0021] Figure 1 Schematically shows an exemplary system architecture to which the search information processing method and apparatus according to the embodiments of the present disclosure can be applied.

[0022] It should be noted that Figure 1The figure shown is only an example of the system architecture to which the embodiments of the present disclosure can be applied, to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios. For example, in another embodiment, the exemplary system architecture to which the search information processing method and apparatus can be applied may include a terminal device, but the terminal device may implement the search information processing method and apparatus provided by the embodiments of the present disclosure without interacting with the server.

[0023] As Figure 1 shown, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 is used as a medium to provide a communication link between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0024] Users may use the terminal devices 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications may be installed on the terminal devices 101, 102, 103, such as knowledge reading applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software, etc. (only as examples).

[0025] The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, and desktop computers, etc.

[0026] The server 105 may be a server providing various services, such as a background management server that supports the content browsed by users using the terminal devices 101, 102, 103 (only as an example). The background management server may analyze and process data such as received user requests, etc., and feedback the processing results (such as web pages, information, or data, etc. obtained or generated according to user requests) to the terminal devices.

[0027] The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, to solve the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services (″Virtual Private Server″, or simply ″VPS″). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0028] It should be noted that the search information processing method provided by the embodiments of the present disclosure can generally be executed by the terminal devices 101, 102, or 103. Correspondingly, the search information processing device provided by the embodiments of the present disclosure can also be set in the terminal devices 101, 102, or 103.

[0029] Alternatively, the search information processing method provided by the embodiments of the present disclosure can generally also be executed by the server 105. Correspondingly, the search information processing device provided by the embodiments of the present disclosure can generally be set in the server 105. The search information processing method provided by the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103, and / or the server 105. Correspondingly, the search information processing device provided by the embodiments of the present disclosure can also be set in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103, and / or the server 105.

[0030] It should be understood that Figure 1 the numbers of the terminal devices, the network, and the server in

[0031] Figure 2 are merely illustrative. According to the implementation requirements, there can be any number of terminal devices, networks, and servers.

[0032] As Figure 2 shown, the search information processing method includes operations S210 to S240.

[0033] In operation S210, in response to obtaining the search information, the large language model is used to process the search information to obtain a plurality of candidate search texts.

[0034] In operation S220, semantic retrieval is performed in a preset search text library based on the candidate search texts to obtain associated texts associated with the candidate search texts. The associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes.

[0035] In operation S230, at least one target search text is determined from the plurality of candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the plurality of candidate search texts.

[0036] In operation S240, the target search text is displayed.

[0037] According to the embodiments of the present disclosure, the search information can include any type of information such as text characters, text words, and characters generated by the target object based on the input operation. The embodiments of the present disclosure do not limit the specific type of the search information.

[0038] According to an embodiment of the present disclosure, the candidate search text may include generated search information associated with the search information. The generated search information may form a text with natural language meaning together with the search information. The generated search information may be associated with the end of the search information. For example, the search information may be "peony", and the candidate search text may be "peony flower purchase", where "flower purchase" may be the generated search information. However, it is not limited to this. The generated search information may also be associated with the beginning of the search information. For example, the search information may be "peony", and the candidate search text may be "how to plant peonies", where "how to plant" may be the generated search information. Or, the generated search information may also be associated with both the beginning and the end of the search information. For example, the search information may be "peony", and the candidate search text may be "how to deal with peony flowers with pests", where "how to deal with" and "flowers with pests" may be the generated search information. The embodiment of the present disclosure does not limit the specific position of the generated search information, and those skilled in the art can select according to actual needs.

[0039] According to an embodiment of the present disclosure, a large language model (LLM: Large Language Model) may include a deep learning model trained using a large amount of text data. The large language model may, for example, be constructed based on a neural network model such as a Transformer model. The large language model can understand the meaning of language texts and generate natural language texts. The large language model can handle natural language tasks. Since the large language model usually contains billions of parameters, the large-scale parameters can help the large language model learn complex patterns in natural language data, so as to perform well in natural language processing (NLP: Natural Language Processing) tasks. Processing the search information based on the large language model can supplement the search information based on the candidate search text, so that the candidate search text can more completely represent the retrieval requirements of the target object.

[0040] According to an embodiment of the present disclosure, processing the search information based on the large language model can understand the typos in the search information based on the semantic understanding ability of the large language model, and make recommendations based on the semantically correct expression, so as to accurately obtain the generated search information.

[0041] According to an embodiment of the present disclosure, the retrieval demand attribute contribution information may characterize the contribution degree of the retrieval result obtained by retrieving based on a preset retrieval text to meeting the retrieval attribute demand. For example, in the case where the retrieval attribute demand characterizes a knowledge retrieval demand, the contribution degree may be characterized by the degree of meeting the knowledge retrieval demand of the target object for the retrieval results recalled by each of the multiple preset retrieval texts. For another example, in the case where the retrieval attribute demand characterizes a transaction demand, the contribution degree may be characterized by the degree of meeting the transaction demand of the target object for the retrieval results recalled by each of the multiple preset retrieval texts.

[0042] It should be noted that the retrieval demand contribution information may be determined based on any method. For example, the retrieval demand contribution information may be determined by the scoring method of the target object for the preset search text. However, it is not limited to this. It may also be determined based on any type of retrieval interaction data such as click-through rate, conversion rate, number of comments, and number of favorites related to the preset search text in the historical time period. The embodiments of the present disclosure do not limit the method for determining the retrieval demand contribution information.

[0043] According to an embodiment of the present disclosure, the associated text may include text that matches the candidate search text, or may also include text with a relatively high semantic similarity to the candidate search text. By determining the associated text related to the candidate search text, a mapping relationship can be established between the candidate search text and the retrieval demand attribute contribution information corresponding to the associated text. In this way, based on the retrieval demand attribute contribution information related to the candidate search text, the contribution degree of the candidate search demand information to the retrieval demand attribute can be evaluated, and the evaluation accuracy of the candidate search demand information can be improved.

[0044] According to an embodiment of the present disclosure, by determining the target search text from multiple candidate search texts according to the retrieval demand attribute contribution information, the obtained target search text can more accurately meet the retrieval demand attribute. Thus, by displaying the target search text, the target object can complete the retrieval operation through the displayed target search text without continuing to input new search information, improving the retrieval efficiency.

[0045] According to an embodiment of the present disclosure, the retrieval demand attribute may include at least one of the following: transaction demand attribute, image demand attribute, text demand attribute.

[0046] According to an embodiment of the present disclosure, the transaction demand attribute may characterize the demand of the target object to conduct transaction behaviors such as product purchase and product lease through search text retrieval. The retrieval demand attribute contribution information corresponding to the transaction demand attribute may characterize the contribution information related to transaction attributes such as transaction amount, transaction conversion rate, and transaction satisfaction.

[0047] According to an embodiment of the present disclosure, the image demand attribute may characterize the retrieval demand for various types of images such as pictures and videos. The retrieval demand attribute contribution information corresponding to the image demand attribute may characterize the contribution information related to image retrieval demand attributes such as image quality and image popularity.

[0048] According to an embodiment of the present disclosure, the text demand attribute may characterize the retrieval demand for text information such as knowledge information and regulatory documents by the search text. The retrieval demand attribute contribution information corresponding to the text demand attribute may characterize the contribution information related to text retrieval demand attributes such as text quality and text popularity.

[0049] It should be noted that the retrieval demand attribute contribution information may be characterized based on a numerical value, but is not limited thereto. It may also be characterized based on a contribution level identifier. The embodiments of the present disclosure do not limit the specific characterization method of the retrieval demand attribute contribution information, and those skilled in the art may select according to actual needs.

[0050] According to an embodiment of the present disclosure, the retrieval demand attribute may further include the demand attribute for retrieving other types of information, and those skilled in the art may select the retrieval demand attribute according to actual needs.

[0051] According to an embodiment of the present disclosure, the search information processing method may further include: in response to obtaining the audio information input by the target object, performing text conversion on the audio information to obtain search information.

[0052] According to an embodiment of the present disclosure, under the condition of obtaining the authorization of the target object, the audio information input by the target object may be collected based on the audio collection device of a terminal device such as a smart phone, and the audio information may be speech recognized based on the retrieval operation of the target object to obtain search information represented in natural language.

[0053] According to an embodiment of the present disclosure, the search information processing method may further include: in response to obtaining the image information input by the target object, performing text recognition on the image information to obtain search information.

[0054] In an example of the present disclosure, under the condition of obtaining the copyright of the image information by the target object, the image information input by the target object may be text recognized based on OCR (Optical Character Recognition) to obtain search information.

[0055] According to an embodiment of the present disclosure, using a large language model to process the search information to obtain multiple candidate search texts may include: using the large language model to process the search information and the historical search information related to the search information to obtain multiple candidate search texts.

[0056] According to an embodiment of the present disclosure, the historical search information may include the search information input by the target object in the historical time period before the search information is input, or may also include search information such as text and images corresponding to the search operations performed by the target object in the historical time period.

[0057] According to an embodiment of the present disclosure, by using a large language model to process the historical search information and the current search information, the upstream search information related to the current search information can be fully learned based on the semantic understanding ability of the large language model, and then the retrieval intention of the target object can be comprehensively understood, so that the output candidate text can more accurately meet the retrieval requirement attributes of the target object, improve the matching degree between the target search text and the retrieval requirement attributes of the target object, and thus improve the retrieval efficiency.

[0058] According to an embodiment of the present disclosure, using a large language model to process the search information and the historical search information related to the search information, obtaining a plurality of candidate search texts includes: updating a preset search prompt template based on the search information and the historical search information to obtain search prompt information; and using the large language model to process the search prompt information to obtain a plurality of candidate search texts.

[0059] According to an embodiment of the present disclosure, the search prompt template (or prompt template) may be information for helping the large language model understand the task of generating a semantic encoding sequence. The search prompt template may be determined based on a prompt token sequence for controlling the accurate prediction of the large language model. The prompt token sequence may include any type of prompt tokens such as characters, fields, and words. The search prompt template may be filled based on the search information and the historical search information to obtain search prompt information.

[0060] In an example of the present disclosure, the search prompt information may be represented based on the following paragraph enclosed by " / / ":

[0061] / / The user's search information is "peony", and the historical search information is "flying insects", "peony price",...... It can be determined that the candidate search texts are: / /

[0062] Using the large language model to process the search prompt information, the obtained plurality of candidate search texts may include: "peony flower flying insect control", "peony flower purchase", "peony flower express delivery", and so on.

[0063] According to an embodiment of the present disclosure, the large language model may be obtained after pre-training based on training samples.

[0064] In one example of the present disclosure, training samples can be constructed based on the sample search text and sample search interaction information collected in a historical time period. The sample search interaction information can include search interaction information such as search text click information, search text replacement rate, conversion rate, etc. for the sample search text. By using the training samples to train an initial large language model, fine-tuning of the generative large language model is achieved, enabling the pre-trained large language model to be applicable to performing specific tasks of predicting candidate search texts based on search information.

[0065] According to an embodiment of the present disclosure, semantic retrieval based on a candidate search text in a preset search text library to obtain associated text related to the candidate search text may include: performing semantic similarity evaluation on the candidate search text and preset search texts in the preset search text library to obtain a similarity evaluation result; and determining, based on the similarity evaluation result, associated text related to the candidate search text from the preset search text library.

[0066] According to an embodiment of the present disclosure, a deep learning model can be used to perform similarity evaluation on the preset search text and the candidate search text. For example, semantic features of the preset search text and the candidate search text can be extracted respectively based on a pre-trained encoder to obtain a preset search text feature and a candidate search text feature. By calculating the similarity between the preset search text feature and the candidate search text feature, a similarity evaluation result is obtained.

[0067] According to an embodiment of the present disclosure, determining, based on the similarity evaluation result, associated text related to the candidate search text from the preset search text library may include determining, as the associated text, the preset search text in the preset search text library that has the highest similarity to the candidate search text and whose similarity evaluation result is greater than a preset similarity threshold. By determining the associated text corresponding to the candidate search text, it is possible to determine the contribution information of the retrieval requirement attributes corresponding to the candidate search text through semantic similarity, thereby enabling an accurate evaluation of the contribution degree of the candidate search text to the retrieval requirement attributes, and further improving the matching degree between the subsequent obtained target search text and the retrieval requirement attributes of the target object, and enhancing the retrieval efficiency.

[0068] According to an embodiment of the present disclosure, semantic retrieval based on a candidate search text in a preset search text library to obtain associated text related to the candidate search text further includes: matching the candidate search text with preset search texts in the preset search text library, and determining the preset search text that matches the candidate search text as the associated text.

[0069] According to an embodiment of the present disclosure, the candidate search text can be fully field-matched with the preset search text, or the candidate search text can also be matched with the preset search text by keyword matching. The embodiment of the present disclosure does not limit the specific matching method.

[0070] In an example of the present disclosure, a preset search text library can be constructed based on the historical search texts that have been input by multiple target objects obtained during a historical time period (for example, within a month). The preset search texts in the preset search text library are determined as high-frequency search texts with a relatively high search frequency, and for each preset rule, the corresponding... When the candidate search text matches the preset search text, it can indicate that the candidate search text output by the large language model is a high-frequency search text, and thus the retrieval requirement attribute contribution information associated with the high-frequency search text can be directly determined.

[0071] In an example of the present disclosure, the candidate search text can first be matched with the preset search texts in the preset search text library to facilitate the determination of the high-frequency search texts in the candidate search text. When the matching result indicates a mismatch, the candidate search text can be evaluated for semantic similarity with the preset search texts in the preset search text library to obtain a similarity evaluation result, and the associated text corresponding to the candidate search text can be determined based on the similarity evaluation result, so as to establish the association relationship between the candidate search text in the low-frequency search scenario and the retrieval requirement attribute contribution information, facilitate the determination of the contribution degree of the low-frequency candidate search text to the retrieval requirement attribute, and avoid missing the candidate search text due to the inability to determine the contribution degree of the candidate search text to the retrieval requirement attribute.

[0072] Figure 3 Schematically shows an application scenario diagram of the search information processing method according to an embodiment of the present disclosure.

[0073] As Figure 3As shown, the search information 311 input by the target object in this application scenario can be the text "peony". Inputting the search information 311 into the large language model 300 can output multiple candidate search texts 321, 322, 323,...... up to 32N. The candidate search texts 321, 322, 323...... up to 32N can be "peony flower delivery", "peony flower purchase", "what is peony"...... up to "where does peony flower grow" respectively. Matching the candidate search texts 321 and 322 with the preset search texts in the preset search text library 330 can obtain a matching result representing the match. The preset search texts in the preset search text library 330 that match the candidate search texts 321 and 322 respectively can be determined as the associated texts associated with the candidate search texts 321 and 322, so as to obtain the retrieval requirement attribute contribution information associated with the candidate search texts 321 and 322 respectively. Performing a semantic similarity evaluation on the candidate search texts 323 and 32N and the preset search texts in the preset search text library 330 to obtain a similarity evaluation result, and then determining the associated texts corresponding to the candidate search texts 323 and 32N respectively, and then determining the retrieval requirement attribute contribution information corresponding to the candidate search texts 323 and 32N respectively.

[0074] According to an embodiment of the present disclosure, the retrieval requirement attribute contribution information is associated with a contribution weight, and the retrieval requirement attribute contribution information can include multiple ones.

[0075] According to an embodiment of the present disclosure, in the case where the retrieval requirement attribute contribution information includes multiple ones, each of the multiple retrieval requirement attribute contribution information can be associated with a contribution weight. The contribution weight can represent the preference degree of the retrieval requirement attribute for the target object among multiple retrieval requirement attributes, or can also represent the importance degree of each of the multiple retrieval requirement attributes. Those skilled in the art can determine the contribution weight based on actual needs, or can also determine multiple contribution weights based on the parameter setting operation of the target object.

[0076] In one example, the multiple retrieval requirement attribute contribution information can include the retrieval requirement attribute contribution information corresponding to the transaction requirement attribute and the image requirement attribute respectively. The target object can make the obtained target search text closer to meeting the transaction requirement of the target object by setting the contribution weight corresponding to the transaction requirement attribute to be greater than the contribution weight corresponding to the image requirement attribute, so as to improve indicators such as the click-through rate and conversion rate of the target search text.

[0077] According to an embodiment of the present disclosure, determining at least one target search text from multiple candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts may include: determining object requirement contribution information related to a candidate search text according to multiple retrieval requirement attribute contribution information associated with the candidate search text and contribution weights corresponding to the multiple retrieval requirement attribute contribution information; and determining a target search text from the multiple candidate search texts according to the object requirement contribution information related to each of the multiple candidate search texts.

[0078] According to an embodiment of the present disclosure, the object requirement contribution information may characterize the contribution degree of a candidate retrieval requirement text to meeting the overall retrieval requirement of a target object.

[0079] According to an embodiment of the present disclosure, the contribution weight and the retrieval requirement attribute contribution information may be represented numerically. For the same retrieval requirement attribute, the contribution value corresponding to the retrieval requirement attribute may be determined by calculating the product between the associated retrieval requirement attribute contribution information and the contribution weight, and the object requirement contribution information related to the candidate search text may be determined by calculating the sum of multiple contribution values. In this way, the contribution degrees of the multiple candidate search texts to meeting the retrieval requirement of the target object can be determined according to the object requirement contribution information related to each of the multiple candidate search texts, so that when the target object does not input a complete retrieval text, the obtained target search text can more accurately meet the overall retrieval requirement of the target object, facilitating the target object to perform retrieval according to the displayed target search text and improving the retrieval efficiency and accuracy.

[0080] According to an embodiment of the present disclosure, the search information processing method may further include: determining object retrieval requirement attributes related to the target object according to the search information and historical interaction information of the target object obtained in relation to the search information; and determining contribution weights corresponding to the multiple retrieval requirement attribute contribution information based on the object retrieval requirement attributes.

[0081] According to an embodiment of the present disclosure, the historical interaction information may include information generated based on interaction operations performed by the target object in a historical time period, and the interaction operations may include, for example, comments, forwards, favorites, searches, etc.

[0082] According to an embodiment of the present disclosure, search information and historical interaction information can be processed based on a pre-trained deep learning model to obtain object retrieval requirement attributes, so that the retrieval requirement attribute preferences of the target object can be further evaluated through the obtained object retrieval requirement attributes. Therefore, by determining the contribution weights according to the obtained object retrieval requirement attributes, it is possible to more accurately represent the retrieval requirement attribute preferences of the target object through the contribution weights, and it is possible to determine a matching target search text according to the retrieval requirement attribute preferences of the target object, further meeting the retrieval requirements of the target object and improving the retrieval efficiency.

[0083] It should be noted that the information acquisition in any embodiment of the present disclosure, including but not limited to the acquisition of historical interaction information, historical search information, etc., is obtained under the condition of obtaining the authorization of the target object. The obtained information has been subjected to confidentiality processing such as desensitization processing and encryption processing, hiding the identity information and avoiding information leakage.

[0084] According to an embodiment of the present disclosure, there can be multiple target search texts, and the multiple target search texts can be respectively associated with object demand contribution information.

[0085] According to an embodiment of the present disclosure, presenting the target search text can include: arranging the multiple target search texts in the interaction interface according to the object demand contribution information related to each of the multiple target search texts.

[0086] According to an embodiment of the present disclosure, the multiple target search texts can be arranged in descending order of the contribution degree represented by the object demand contribution information in the interaction interface, so as to facilitate the target object to select a target search text to perform a retrieval operation.

[0087] According to an embodiment of the present disclosure, the search information processing method can further include: in response to a search operation for the target search text, performing a page jump operation, and the page jump operation indicates a jump to a page corresponding to the search operation.

[0088] Figure 4 Schematically shows an application scenario diagram of the search information processing method according to another embodiment of the present disclosure.

[0089] As Figure 4As shown, the application scenario may include an interaction interface 400, and the interaction interface 400 may include a search box 410. The target object may enter the search information "peony" in the retrieval box 410. According to the search information processing method provided by the embodiments of the present disclosure, multiple target search texts can be obtained and displayed in the search recommendation area 420. The multiple target search texts may include "peony flower mailing", "peony flower purchase", "peony flower seed purchase"... "peony flower cultivation". The target object can perform a click operation on any one of the target search texts to conduct a search, so as to avoid having to continue entering the complete search text, thereby improving the search efficiency.

[0090] According to the embodiments of the present disclosure, the search operation for the search text may include a click operation, but is not limited thereto, and may also include any type of operation such as a double-click operation, a drag operation, a text editing operation, etc. The embodiments of the present disclosure do not limit the specific operation type of the search operation, and those skilled in the art can select according to actual needs.

[0091] According to the embodiments of the present disclosure, by performing a page jump operation according to the search operation of the target object for the target search text, the page obtained after the jump can be quickly displayed to the target object to timely meet the retrieval needs of the target object.

[0092] Figure 5 A block diagram of a search information processing device according to an embodiment of the present disclosure is schematically shown.

[0093] As Figure 5 shown, the search information processing device 500 may include: a candidate search text obtaining module 510, an associated text obtaining module 520, a target search text determining module 530, and a display module 540.

[0094] The candidate search text obtaining module 510 is configured to, in response to obtaining the search information, process the search information using a large language model to obtain a plurality of candidate search texts, where the candidate search texts include generated search information associated with the search information.

[0095] The associated text obtaining module 520 is configured to perform semantic retrieval in a preset search text library based on the candidate search texts to obtain associated texts associated with the candidate search texts, where the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes.

[0096] The target search text determining module 530 is configured to determine at least one target search text from the plurality of candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the plurality of candidate search texts.

[0097] A display module 540 for displaying a target search text.

[0098] According to an embodiment of the present disclosure, the retrieval requirement attribute contribution information is associated with a contribution weight, and the retrieval requirement attribute contribution information includes multiple ones.

[0099] According to an embodiment of the present disclosure, the target search text determination module includes: an object requirement contribution information determination sub-module and a target search text determination sub-module.

[0100] The object requirement contribution information determination sub-module is configured to determine object requirement contribution information related to a candidate search text according to multiple retrieval requirement attribute contribution information associated with the candidate search text, and contribution weights corresponding to the multiple retrieval requirement attribute contribution information respectively.

[0101] The target search text determination sub-module is configured to determine a target search text from multiple candidate search texts according to the object requirement contribution information related to each of the multiple candidate search texts.

[0102] According to an embodiment of the present disclosure, the search information processing device further includes: an object retrieval requirement attribute determination module and a contribution weight determination module.

[0103] The object retrieval requirement attribute determination module is configured to determine object retrieval requirement attributes related to a target object according to search information and historical interaction information of the target object related to the search information obtained.

[0104] The contribution weight determination module is configured to determine contribution weights corresponding to the multiple retrieval requirement attribute contribution information respectively based on the object retrieval requirement attributes.

[0105] According to an embodiment of the present disclosure, the target search text includes multiple ones.

[0106] According to an embodiment of the present disclosure, the display module includes a display sub-module.

[0107] The display sub-module is configured to arrange multiple target search texts in an interaction interface according to the object requirement contribution information related to each of the multiple target search texts.

[0108] According to an embodiment of the present disclosure, the candidate search text obtaining module includes a candidate search text obtaining sub-module.

[0109] The candidate search text obtaining sub-module is configured to use a large language model to process search information and historical search information related to the search information to obtain multiple candidate search texts.

[0110] According to an embodiment of the present disclosure, the candidate search text obtaining sub-module includes: a search prompt information obtaining unit and a candidate search text obtaining unit.

[0111] A search hint information acquisition unit, configured to update a preset search hint template based on search information and historical search information to obtain search hint information.

[0112] A candidate search text acquisition unit, configured to process the search hint information by using a large language model to obtain multiple candidate search texts.

[0113] According to an embodiment of the present disclosure, the associated text acquisition module includes a similarity evaluation sub-module and a first associated text determination sub-module.

[0114] The similarity evaluation sub-module is configured to perform semantic similarity evaluation on the candidate search text and a preset search text in a preset search text library to obtain a similarity evaluation result.

[0115] The first associated text determination sub-module is configured to determine an associated text associated with the candidate search text from the preset search text library according to the similarity evaluation result.

[0116] According to an embodiment of the present disclosure, the associated text acquisition module further includes a second associated text determination sub-module.

[0117] The second associated text determination sub-module is configured to match the candidate search text with a preset search text in the preset search text library, and determine the preset search text that matches the candidate search text as the associated text.

[0118] According to an embodiment of the present disclosure, the retrieval requirement attribute includes at least one of the following: a transaction requirement attribute, an image requirement attribute, and a text requirement attribute.

[0119] According to an embodiment of the present disclosure, the search information processing device further includes a page jump operation execution module.

[0120] The page jump operation execution module is configured to execute a page jump operation in response to a search operation for a target search text, and the page jump operation instructs to jump to a page corresponding to the search operation.

[0121] According to an embodiment of the present disclosure, the search information processing device further includes a first search information acquisition module.

[0122] The first search information acquisition module is configured to perform text conversion on audio information in response to obtaining audio information input by a target object to obtain search information.

[0123] According to an embodiment of the present disclosure, the search information processing device further includes a second search information acquisition module.

[0124] The second search information acquisition module is configured to perform text recognition on image information in response to obtaining image information input by a target object to obtain search information.

[0125] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0126] According to an embodiment of the present disclosure, an electronic device includes: 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 to enable the at least one processor to execute the method as described above.

[0127] According to an embodiment of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method as described above.

[0128] According to an embodiment of the present disclosure, a computer program product includes a computer program, and the computer program implements the method as described above when executed by a processor.

[0129] Figure 6 A block diagram of an electronic device suitable for implementing a search information processing method according to an embodiment of the present disclosure is schematically shown. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0130] As Figure 6 shown, the device 600 includes a computing unit 601, which can execute various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0131] Multiple components in device 600 are connected to I / O interface 605, including: input unit 606, such as a keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as a disk, optical disc, etc.; and communication unit 609, such as a network card, modem, wireless communication transceiver, etc. Communication unit 609 allows device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0132] Computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit 601 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, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Computing unit 601 executes the various methods and processes described above, such as the search information processing method. For example, in some embodiments, the search information processing method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by computing unit 601, one or more steps of the search information processing method described above can be executed. Alternatively, in other embodiments, computing unit 601 can be configured to execute the search information processing method in any other suitable manner (e.g., by means of firmware).

[0133] Various embodiments of the systems and techniques described above in this article 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-a-chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments 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, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0134] The program code for implementing the methods 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, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0135] In the context of the present disclosure, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. 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, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0136] In order to provide interaction with a user, the systems and techniques described herein may 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 a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0137] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected with each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0138] A computer system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

[0139] It should be understood that various forms of the processes shown above can be used, steps can be reordered, added, or deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitation is imposed herein.

[0140] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for processing search information, comprising: In response to obtaining the search information, updating a preset search prompt template based on the search information and historical search information related to the search information to obtain search prompt information; Processing the search prompt information by using a large language model to obtain multiple candidate search texts representing the retrieval requirement attributes of the target object, wherein the candidate search texts include generated search information associated with the search information, and the retrieval requirement attributes include at least one of the following: transaction requirement attribute, image requirement attribute, text requirement attribute; Performing semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts, wherein the associated texts are associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attributes; Determining at least one target search text from the multiple candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts, and the retrieval requirement attribute contribution information corresponding to the candidate search text is determined by the retrieval requirement attribute contribution information of the associated text; and Displaying the target search text.

2. The method according to claim 1, wherein, the retrieval requirement attribute contribution information is associated with contribution weights, and there are multiple pieces of the retrieval requirement attribute contribution information, wherein the determining at least one target search text from the multiple candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the multiple candidate search texts includes: Determining object requirement contribution information related to the candidate search text according to multiple pieces of retrieval requirement attribute contribution information associated with the candidate search text and the contribution weights corresponding to each of the multiple pieces of retrieval requirement attribute contribution information; and Determining the target search text from the multiple candidate search texts according to the object requirement contribution information related to each of the multiple candidate search texts.

3. The method according to claim 2, further comprising: Determining object retrieval requirement attributes related to the target object according to the search information and historical interaction information of the target object related to the search information obtained; and Based on the object retrieval requirement attributes, determining the contribution weights corresponding to each of the multiple pieces of retrieval requirement attribute contribution information.

4. The method according to claim 2, wherein, the target search texts include multiple; wherein the displaying the target search text includes: Arranging the multiple target search texts in an interaction interface according to the object requirement contribution information related to each of the multiple target search texts.

5. The method according to claim 1, wherein, the performing semantic retrieval on the candidate search texts in a preset search text library to obtain associated texts associated with the candidate search texts includes: Performing semantic similarity evaluation on the candidate search texts and preset search texts in the preset search text library to obtain a similarity evaluation result; Determine associated text associated with the candidate search text from the preset search text library according to the similarity evaluation result.

6. The method according to claim 5, wherein the semantic retrieval based on the candidate search text in the preset search text library to obtain associated text associated with the candidate search text further includes: matching the candidate search text with the preset search text in the preset search text library, and determining the preset search text that matches the candidate search text as the associated text.

7. The method according to any one of claims 1 to 6, further comprising: In response to a search operation for the target search text, perform a page jump operation, and the page jump operation indicates a jump to a page corresponding to the search operation.

8. The method according to any one of claims 1 to 6, further comprising: In response to obtaining audio information input by a target object, perform text conversion on the audio information to obtain the search information.

9. The method according to any one of claims 1 to 6, further comprising: In response to obtaining image information input by a target object, perform text recognition on the image information to obtain the search information.

10. A search information processing device, comprising: A candidate search text obtaining module, configured to, in response to obtaining search information, process the search information using a large language model to obtain a plurality of candidate search texts, wherein the candidate search texts include generated search information associated with the search information; An associated text obtaining module, configured to perform semantic retrieval on the candidate search text in a preset search text library to obtain associated text associated with the candidate search text, wherein the associated text is associated with retrieval requirement attribute contribution information, and the retrieval requirement attribute contribution information represents the contribution degree to the retrieval requirement attribute, and the retrieval requirement attribute includes at least one of the following: transaction requirement attribute, image requirement attribute, text requirement attribute; A target search text determining module, configured to determine at least one target search text from the plurality of candidate search texts according to the retrieval requirement attribute contribution information corresponding to each of the plurality of candidate search texts, and the retrieval requirement attribute contribution information corresponding to the candidate search text is determined by the retrieval requirement attribute contribution information of the associated text; and A display module, configured to display the target search text; wherein, the candidate search text obtaining module includes: A candidate search text obtaining sub-module, configured to process the search information and historical search information related to the search information using the large language model to obtain the plurality of candidate search texts representing the retrieval requirement attributes of the target object; wherein, the candidate search text obtaining sub-module includes: A search prompt information obtaining unit, configured to update a preset search prompt template based on the search information and the historical search information to obtain search prompt information; and A candidate search text obtaining unit, configured to process the search prompt information using the large language model to obtain the plurality of candidate search texts, and the candidate search texts represent the retrieval requirement attributes of the target object.

11. The apparatus according to claim 10, wherein, the retrieval requirement attribute contribution information is associated with contribution weights, and there are multiple pieces of the retrieval requirement attribute contribution information, wherein, the target search text determination module includes: an object requirement contribution information determination sub-module, configured to determine object requirement contribution information related to the candidate search text according to multiple pieces of retrieval requirement attribute contribution information associated with the candidate search text, and contribution weights corresponding to the multiple pieces of retrieval requirement attribute contribution information respectively; and a target search text determination sub-module, configured to determine the target search text from multiple candidate search texts according to the object requirement contribution information related to each of the multiple candidate search texts.

12. The apparatus according to claim 11, further including: an object retrieval requirement attribute determination module, configured to determine object retrieval requirement attributes related to the target object according to the search information and historical interaction information of the target object related to the search information obtained; and a contribution weight determination module, configured to determine contribution weights corresponding to the multiple pieces of retrieval requirement attribute contribution information respectively based on the object retrieval requirement attributes.

13. The apparatus according to claim 11, wherein, there are multiple target search texts; wherein, the display module includes: a display sub-module, configured to arrange the multiple target search texts in an interaction interface according to the object requirement contribution information related to each of the multiple target search texts.

14. The apparatus according to claim 10, wherein, the associated text obtaining module includes: a similarity evaluation sub-module, configured to perform semantic similarity evaluation on the candidate search text and a preset search text in a preset search text library to obtain a similarity evaluation result; a first associated text determination sub-module, configured to determine an associated text related to the candidate search text from the preset search text library according to the similarity evaluation result.

15. The apparatus according to claim 14, wherein, the associated text obtaining module further includes: a second associated text determination sub-module, configured to match the candidate search text with a preset search text in the preset search text library, and determine the preset search text that matches the candidate search text as the associated text.

16. The apparatus according to any one of claims 10 to 15, further including: a page jump operation execution module, configured to execute a page jump operation in response to a search operation on the target search text, and the page jump operation instructs to jump to a page corresponding to the search operation.

17. The apparatus according to any one of claims 10 to 15, further including: a first search information obtaining module, configured to perform text conversion on audio information input by a target object in response to obtaining the audio information, to obtain the search information.

18. The apparatus according to any one of claims 10 to 15, further including: a second search information obtaining module, configured to perform text recognition on image information input by a target object in response to obtaining the image information, to obtain the search information.

19. An electronic device, 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 to enable the at least one processor to execute the method according to any one of claims 1 to 9.

20. A non-transitory computer-readable storage medium storing computer instructions, wherein, the computer instructions are for causing a computer to execute the method according to any one of claims 1 to 9.

21. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 9.

Images