Commodity searching method and electronic equipment

Abstract

The embodiment of the invention discloses a commodity searching method and electronic equipment, and the method comprises the steps that a commodity searching interaction interface is provided, the interface comprises a first area and a second area, the first area is used for providing an artificial intelligence AI interaction component so as to receive a search demand expressed by a user in a mode of inputting a natural language statement, and the second area is used for providing a second area; through multiple rounds of interaction with an AI large model, receiving a statement which is input by the user under the heuristic of the reply content of the AI large model and is used for clearly expressing the search demand of the user; the second area is used for providing a plurality of attribute options for defining search requirements and a plurality of attribute value options corresponding to the attribute options according to interaction conditions in the first area so as to receive search requirement information expressed by the user in a mode of selecting attributes and attribute values; and in each round of interaction, carrying out information fusion on the input statements collected through the first region and the attributes and attribute value selection results collected through the second region so as to provide corresponding commodity search results according to information fusion results. Through the embodiment of the invention, the user can be helped to express the search demand, and the search efficiency is improved.

Description

Technical Field

[0001] The present application relates to the technical field of commodity information search, and in particular to a commodity search method and electronic equipment. Background Art

[0002] In a product information service system (also known as an e-commerce platform), product search is an important way for users to obtain product information. Users can initiate product searches by entering keywords, and the system can return a list of products that match the keywords, from which users can select the products they are interested in, and so on.

[0003] In the traditional way, buyers are limited to searching through short query words in the search box, which may lead to low matching efficiency or fail to meet detailed and accurate search needs. In addition, when users give feedback and think about the search results to perform the next step of search word optimization, they may face the burden of word choice and sentence construction. Summary of the invention

[0004] The present application provides a product search method and an electronic device, which can help users express their search needs and improve search efficiency.

[0005] This application provides the following solutions:

[0006] A commodity search method, comprising:

[0007] A commodity search interactive interface is provided, the interface including a first area and a second area, the first area is used to provide an artificial intelligence (AI) interactive component to receive search requirements expressed by a user by inputting natural language sentences, and to receive sentences input by the user inspired by the reply content of the AI ​​big model to clarify the search requirements through multiple rounds of interaction with the AI ​​big model; the second area is used to provide multiple attribute options for clarifying the search requirements and multiple attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search requirement information expressed by the user by selecting the attributes and attribute values;

[0008] Product search results are provided based on the information collected through the first area and the second area. The product search results are generated based on the information fusion result after the input sentence collected through the first area and the attribute and attribute value selection results collected through the second area are fused.

[0009] Among them, it also includes:

[0010] After the attributes and attribute value selection results are collected through the second area, the attributes and attribute value selection results are rewritten into sentences expressed in natural language and displayed in the dialogue area of ​​the first area.

[0011] Wherein, when performing the information fusion, it includes:

[0012] The multiple rounds of input statements collected through the first area are rewritten into structured search query statements suitable for search engines, and are merged with the attribute and attribute value selection results collected through the second area to generate a complete structured search query statement, so as to generate corresponding search results according to the structured search query statement.

[0013] Among them, when generating the complete structured search query statement, the AI ​​big model is used to perform natural language understanding on the multiple rounds of input statements and the attribute and attribute value selection results, so that based on understanding the user's search needs, the multiple rounds of input statements are rewritten and the complete structured search query statement is generated.

[0014] Among them, it also includes:

[0015] After obtaining the search results, the AI ​​big model is used to analyze the match between the search results and the results of understanding the user's search needs, and the feedback content in the first area and / or the attribute options and attribute value options displayed in the second area are updated.

[0016] The user includes merchant buyer users, and the merchant buyer users include: buyer users who purchase or customize commodities in bulk through the current commodity information service system for resale or production purposes;

[0017] The search requirements include: the requirement for group purchase of multiple different minimum stock keeping units (SKUs) of the same product; wherein the SKUs are expressed by combining attribute values ​​in multiple dimensions;

[0018] In the process of collecting search requirements, multiple selections of different attribute values ​​under the same attribute are supported.

[0019] Among them, when generating the complete structured search query statement, the selected multiple different attribute values ​​under the same attribute are set to a logical OR relationship, so that when providing corresponding product search results, according to the logical OR relationship, the products that can meet the user's purchasing needs in terms of corresponding SKU supply are screened from the product library and displayed preferentially in the product search results.

[0020] A commodity search method, comprising:

[0021] Receive a product search request submitted by a client side, wherein the product search request is generated after the client collects user expression information on search needs through a product search interactive interface, wherein the interface includes a first area and a second area, wherein the first area is used to provide an AI interaction component to receive the search needs expressed by the user by inputting natural language sentences, and receive, through multiple rounds of interaction with the AI ​​big model, sentences input by the user inspired by the reply content of the AI ​​big model to clarify the search needs; the second area is used to provide multiple attribute options for clarifying the search needs and multiple attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search demand information expressed by the user by selecting the attributes and attribute values;

[0022] Performing information fusion on the input sentences collected through the first area and the attributes and attribute value selection results collected through the second area;

[0023] Generate product search results based on the information fusion results and return them to the client for display.

[0024] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any of the methods described above.

[0025] An electronic device, comprising:

[0026] one or more processors; and

[0027] A memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of any of the methods described above.

[0028] A computer program product comprises a computer program / computer executable instructions, wherein the computer program / computer executable instructions implement the steps of any of the aforementioned methods when executed by a processor in an electronic device.

[0029] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0030] Through the embodiments of the present application, since a first area (LUI, natural language user interface) and a second area (GUI, graphical user interface) can be provided in the search interaction interface, the user can input natural language sentences through the LUI and select attributes and attribute values ​​through the GUI during the search process. Therefore, the flexibility of the LUI can be used to give users the opportunity to express their search needs in a fine-grained and delicate manner, and to clarify, refine, and clarify their own search needs during multiple rounds of dialogue; at the same time, the GUI is used to provide visual search demand options and provide candidates for user needs, so that users can quickly and efficiently adjust and refine search conditions. This design is conducive to improving user experience, reducing the possibility of input errors, and enhancing matching efficiency. Afterwards, the information collected from the LUI and GUI can be integrated and the corresponding search results can be given.

[0031] Of course, any product implementing the present application does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0033] Figure 1 It is a schematic diagram of the system architecture provided by the embodiment of the present application;

[0034] Figure 2 is a flowchart of a client method provided by an embodiment of the present application;

[0035] Figure 3 is a schematic diagram of an interface provided by an embodiment of the present application;

[0036] Figure 4 is another schematic diagram of the interface provided in the embodiment of the present application;

[0037] Figure 5 is a flow chart of the server method provided in an embodiment of the present application;

[0038] Figure 6 It is a schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0040] In order to facilitate understanding of the solution provided by the embodiment of the present application, it should be first explained that in the existing commodity search system, there are also some solutions that help users to more finely or accurately express their search needs. For example, in one solution, considering that after a user enters a keyword to initiate a search, there is often a need to narrow the search scope. For example, the user enters "dress" to initiate a search, but because there are many commodities under the "dress" category, the user may need to narrow the search scope, for example, modify the keyword to "French court style dress". For this reason, some systems can provide users with "filter" or "screening" functions, which can include filtering by category, etc. That is, after the user enters "dress" to initiate a search, the system can provide a "filter" for filtering by category, etc., which can provide multiple different options including "French court style dress" for the user to choose, thereby avoiding the user from re-entering keywords. However, the above-mentioned filter may involve operations such as step-by-step selection of categories during use, which has operational difficulty and usability issues, and the usage rate is not high.

[0041] With the emergence of AI (Artificial Intelligence) models, especially AI large-scale parameter models (AI large models for short), they have been applied in multiple scenarios and also provide a new idea for product search. In fact, some systems have also made related attempts. In their specific solutions, users can express their search needs by entering long natural language sentences instead of short keywords, such as "I want to buy a smart watch". After that, AI assistants can help users narrow the search scope. For example, AI assistants can display some options for users to choose, including: buying for men, women or children, what is the approximate price range, whether fitness tracking functions are needed, whether there are high requirements for intelligence, etc. Users can make choices for these options respectively, and then give corresponding product search results.

[0042] Compared with the method of providing filters through the sidebar, the above scheme is more conducive to obtaining more accurate search results and is more convenient and easy to use. However, the inventor of the present application found in the process of implementing the present application that the above scheme still has at least the following problems: the above scheme is mainly designed for the shopping scenarios of Class C buyers. The so-called Class C buyers refer to individual consumers or terminal buyers. When purchasing goods, such Class C buyers mainly use them for personal consumption. Therefore, the required decision factors are usually relatively few, and under each specific decision factor, the specific parameter values ​​are usually deterministic. Usually, only one simple conversation is needed to clarify the user's specific search needs, and then the search results can be given. Therefore, the above AI-based search system in the prior art usually only supports a single round of conversation.

[0043] However, in actual applications, there may be some more complicated situations. For example, some users may not be able to express the required product categories very specifically before initiating a search, but can only express their needs relatively abstractly. For example, the user may only express the need to attend a certain occasion, in which there is no information about the required product categories, and there is no attribute value on a specific attribute, etc. Therefore, it is impossible to directly initiate a search, and it is usually difficult to accurately express their specific search needs through a round of dialogue. In addition, the buyer users in some systems may be Class B buyers, that is, this type of buyer usually needs to purchase or customize goods in bulk for resale or production purposes. For these buyers, their search needs may be more complicated, and a simple single-round dialogue is more difficult to fully and accurately express their search needs. Correspondingly, it is difficult to provide accurate search results.

[0044] In view of the above situation, the inventor of the present application found in the process of realizing the present application that in order to guide users to express their search needs more completely and accurately, the existing single-round dialogue can be changed to a multi-round dialogue, and the user can be gradually inspired to express more accurate search needs in the process of multi-round dialogue. However, since the user needs to perform multiple input operations, including voice input or keyboard input, etc., during the multi-round dialogue with the AI ​​assistant, there may be complex operations and low interaction efficiency, especially in the scenario of Class B buyers. Since the search needs are more complex (for example, in the scenario of purchasing or replenishing, Class B buyers may need to purchase different specifications and colors of the same product, etc.), therefore, if only a multi-round AI interaction function is provided, the user may still need to spend a relatively long time to complete the expression of the search needs and then obtain more accurate search results.

[0045] To this end, in an embodiment of the present application, an implementation scheme for joint interaction based on LUI (Language User Interface) and GUI (Graphical User Interface) is provided, wherein LUI can be an interface for multiple rounds of interaction with AI. In addition, a GUI part can also be provided in an embodiment of the present application, which can provide users with multiple related attributes and optional attribute values ​​during the AI ​​interaction process, so that users can express their search needs by selecting operations. Among them, the above-mentioned LUI and GUI can exist in the same interface, so that users can interact with the AI ​​assistant through voice input or keyboard input in the LUI area to gradually and accurately express their search needs, and can also express their search needs by selecting attributes and attribute values ​​in the GUI area. Correspondingly, the system side can fuse the user input information and selection information collected from the two areas, and give corresponding search results. Among them, in the process of giving search results, the AI ​​large model can also infer the user's search demand information from the fusion information, and rewrite it into a structured query statement that the search engine system can process, so that the search engine system can give search results. In this way, users can gradually and clearly express their search needs with the help of AI assistants. However, there are not only interactive ways to express needs such as voice or keyboard input, but also combined with attribute selection methods to express their search needs through multiple channels. Therefore, it is easy to improve search efficiency, especially when Class B buyers need to express more complex search needs. The advantages of this method in improving efficiency will be more obvious.

[0046] From the perspective of system architecture, the embodiment of the present application can provide a product search interaction method based on the combination of LUI and GUI. This solution can be implemented in the product information service system. Figure 1, the system may include a client and a server, wherein the client may exist in the form of a Web page, an H5 page or a mobile App (application), etc., for providing a display of a specific interface, including the display of the LUI and the GUI, and interacting with the user; wherein the LUI part may provide an input box, and display the user input content and the reply content of the AI ​​assistant generated during the dialogue between the two parties, and the GUI part may display multiple attribute options, each of which may include multiple attribute value options, and these attributes and attribute values ​​are related to the dialogue content of the LUI part. The server is used to provide specific data support, including the call of the AI ​​big model, and the provision of relevant search results, etc. Specifically, the server can call the AI ​​big model to summarize the multiple rounds of contextual content collected by the LUI part, and can also perform structured rewriting; in addition, the rewriting results can be integrated with the structured demand information collected from the GUI, and a structured query statement suitable for the search engine can be generated and provided to the search engine for search. In an optional manner, content updates can also be performed based on search results and demand understanding results, including updates to the reply content in the LUI, updates to the attributes and attribute value options displayed in the GUI, and so on. It should be noted that when the hardware resources of the terminal device are sufficient to support the corresponding computing power, information fusion, calling of large AI models and other processing can also be completed on the client side.

[0047] The specific implementation scheme provided in the embodiments of the present application is described in detail below.

[0048] Embodiment 1

[0049] First, the first embodiment of the present application provides a commodity search method from the perspective of the client, see Figure 2 , the method may include:

[0050] S201: Provide a product search interaction interface, the interface including a first area and a second area, the first area is used to provide an artificial intelligence (AI) interaction component to receive search requirements expressed by users by inputting natural language sentences, and receive sentences input by the user inspired by the reply content of the AI ​​big model to clarify their search requirements through multiple rounds of interaction with the AI ​​big model; the second area is used to provide multiple attribute options for clarifying search requirements and multiple attribute value options corresponding to the attribute options based on the interaction situation in the first area, so as to receive search requirement information expressed by the user by selecting the attributes and attribute values.

[0051] In the embodiment of the present application, a product search interactive interface can be provided for users. Figure 3, the interface may include a first area 31 and a second area 32. Of course, in the initial display state of the interactive interface, the user can first input a natural language sentence through the input control 33 in the first area (including input controls such as voice or keyboard) to express his search needs. Among them, the natural language sentence input by the user can be a long sentence or other sentence that is more suitable for use in a dialogue scenario, that is, it can no longer be a short keyword, phrase, etc., but can be a sentence including a variety of different sentence components, for example, "Help me find a red plastic cup" (of course, in actual applications, the information initially input by the user may be very vague or abstract, for example, "I need to attend a dinner party and don't know what to wear", etc.), etc. After that, the AI ​​big model can understand the sentence input by the user and guide the user to input more information to make his search needs clearer. For example, the content of the AI ​​big model's reply can be as follows: Figure 3 As shown in 34, it can be specifically:

[0052] "Choosing red plastic cups is a practical and popular choice for various events and parties. When choosing red plastic cups, there are also several aspects to consider:

[0053] Capacity: Make sure the cup is the volume you need, such as 12 ounces, 16 ounces, or 18 ounces;

[0054] Durability: Look for cups made of high-quality materials to avoid damage during use;

[0055] Reusability: Some plastic cups are designed to be reusable, which can be more economical and environmentally friendly;

[0056] If you need to source any other products please let me know.”

[0057] Of course, while the AI ​​big model gives the above reply content, it can also display a variety of optional attributes in the second area 32, as well as a plurality of optional attribute values ​​under each attribute. Among them, which attributes and attribute values ​​are specifically displayed in the second area 32 can be determined according to the user's interaction in the first area 31, and these attributes and attribute value options can also be used to help users clarify their search needs. For example, in the aforementioned example, the sentence input by the user is "Help me find a red plastic cup", then the attributes and attribute values ​​displayed in the second area can also be related to "red plastic cup", for example, including attributes such as category, feature, style, etc., wherein the category attribute can include the following attribute values: plastic cup, water bottle, disposable plastic cup, disposable cup, large cup, etc.; the feature attribute can include the following attribute values: biodegradable, Bpa-free, disposable, environmentally friendly, food grade, reusable, storage, etc.; the style attribute can include attribute values ​​such as cartoon, children, contemporary, local, eclectic, etc.

[0058] In this way, in the next round of interaction, the user can continue to input more specific search requirements through the input control 33 according to the guidance of the AI ​​assistant, or, since the attributes, attribute values, etc. displayed in the second area are also related to the interaction process, the attribute values ​​can be selected through the options displayed in the second area 32. Alternatively, the user can input natural language sentences in the first area 31 and select attribute values ​​in the second area 32, so that the user can express his search needs through a variety of different ways.

[0059] After completing the input of natural language sentences or the selection of attributes and attribute values, users can initiate a specific search by clicking options such as "Apply" to obtain new search results. After that, users can continue to talk to the AI ​​assistant, and the second area will also update specific attributes and attribute value options. Among them, attributes that have been selected for attribute values ​​in the previous round or other rounds will no longer appear in the second area, thereby guiding users to express their search needs through attribute value selection on more attributes. After the user completes the input and / or selection operation, he can initiate the next search until he sees the product that meets his real needs in the search results.

[0060] S202: providing product search results based on the information collected through the first area and the second area, wherein the product search results are generated based on the information fusion result after information fusion of the input sentence collected through the first area and the attribute and attribute value selection results collected through the second area.

[0061] In each round of interaction, since the user may input natural sentences through the first area and select attribute values ​​through the second area, the system can fuse the input sentences collected through the first area and the attributes and attribute value selection results collected through the second area, and then give the corresponding product search results. Of course, in each round of interaction, the information collected in the current round and in multiple rounds of historical interactions can be processed together to jointly infer the user's search needs and give search results.

[0062] Specifically, when information fusion is performed, since the second area provides attribute and attribute value options, after the user selects one or some attribute values, the selection result can naturally carry structured information, for example, it can be a key-value pair format in the form of <attribute, attribute value>, etc. However, for the input information collected through the first area, since the user expresses it in the form of natural language sentences, this information does not belong to structured information. In order to facilitate the subsequent search results, the unstructured input statements collected through the first area can also be rewritten into structured search query statements suitable for search engines. For example, in a specific implementation, some keywords and other information can be extracted from the unstructured input statements, and then rewritten into structured search query statements suitable for the search system, and then fused with the structured information collected through the second area to generate a complete structured search query statement. The structured search query statement can be provided to the search engine, and the search engine executes the specific search logic to give the product search results.

[0063] It should be noted that, in the specific implementation, the user in the embodiment of the present application may be a merchant buyer user, that is, a Class B buyer user. This Class B buyer user refers to a buyer user who purchases or customizes goods in bulk through the current commodity information service system for resale or production purposes. When such a user searches for goods, his specific search needs may be different from those of Class C buyers. Class C buyers usually shop for their own use or as gifts, and usually only purchase single items. They need to select an attribute value in each dimension, and these attribute values ​​can correspond to a SKU (minimum inventory unit) of the product; however, Class B buyers usually purchase in bulk and may need to replenish multiple SKUs. For example, for plastic cups, not only red but also green, yellow, etc. may be required. In terms of capacity, not only 500ml is required, but also 350ml, 600ml, etc. are required. In view of this situation, in the embodiment of the present application, it is possible to support the need to purchase multiple different SKUs of the same product, and in the process of collecting search needs, it is possible to support multiple selections of multiple different attribute values ​​under the same attribute. Of course, such multiple selections may include expressing multiple different attribute values ​​under the same attribute in the natural language sentence input in the first area, and may also include multiple selections of the listed attribute values ​​in the second area, etc. In other words, in view of the common demand of Class B buyers for grouping goods with multiple SKUs, multiple attribute values ​​under one attribute can be conveniently selected by clicking in the GUI. Similarly, the required multiple attribute values ​​can also be expressed verbally in the LUI.

[0064] For the above-mentioned multi-SKU grouping situation, in the process of demand fusion, the logical relationship of OR (logical OR) and AND (logical AND) between multiple attribute values ​​can be accurately retained according to semantic understanding and rule restrictions, and then take effect in the search engine. That is to say, when generating the complete structured search query statement, the selected multiple different attribute values ​​under the same attribute can be set to a logical OR relationship, so that when providing the corresponding product search results, the products that can meet the user's purchasing needs in terms of the corresponding SKU supply can be screened from the product library according to the logical OR relationship, and are displayed preferentially in the product search results. In this way, since the logic of OR can be supported, it can ensure that users can match at the granularity of SPU, rather than only being guided by the SKU perspective that satisfies all attributes.

[0065] In the search engine, the inverted index of products can be optimized to ensure that when recalling keywords, the AND and OR relationships between the multi-attribute values ​​under each attribute structured in the demand fusion module are followed to process the recalled products of each attribute. In addition, the product relevance processing process can be optimized so that products that accurately match the category and attribute requirements are presented to users first.

[0066] It should be noted here that when the user selects attributes and attribute values ​​through the second area, the AI ​​large model can also rewrite the attribute and attribute value selection results into sentences expressed in natural language and display them in the dialogue area in the first area, so as to prompt the user that the attribute and attribute value selection operation made in the second area has the same effect as entering such a natural language sentence in the first area. Since natural language sentences are easier for users to understand, the attribute and attribute value selection results performed by the user through the second area can be rewritten into the form of such natural language sentences. It can also be reflected in the dialogue area of ​​the LUI part in the form of multiple rounds of dialogue, which can enable users to more intuitively understand whether the current search results match their search needs.

[0067] For example, in Figure 3 In the example shown, suppose the user enters "Help me find a red plastic cup" in the first area in the first round. After multiple optional attributes and attribute values ​​are displayed in the second area, the user selects attribute values ​​such as "with lid" and "with handle". After the user clicks "Apply", Figure 4 As shown, in addition to displaying the updated search results, the dialogue content such as "It is better to have a lid and a handle" can also be displayed in the dialogue area (such as Figure 4 As shown in 41, the GUI part is in Figure 4 (not shown). That is to say, what the user actually needs to express in the second round of interaction may be a requirement such as "it is better to have a lid and a handle". However, the user does not need to enter such a sentence by voice or keyboard input, but can click and select the attribute values ​​such as "with a lid" and "with a handle" displayed in the second area. Correspondingly, AI can rewrite this selection result into the above-mentioned sentence "it is better to have a lid and a handle" and give the search result. In this way, since the selection operation of attributes and attribute values ​​is more efficient than the input operation of natural language sentences, the selection results of attributes and attribute values ​​can be rewritten into natural language sentences and integrated into the dialogue in the first area. Therefore, while improving efficiency, the continuity of the dialogue content and the integration of the operation results of the two areas can also be achieved. In addition, the user can also check whether the system understands the user's needs correctly through the rewritten natural language sentences. If not, it can make timely adjustments in the next round of interaction, etc., thereby improving the user experience.

[0068] It should also be noted that in actual applications, it is not necessary to construct a structured search query statement, but to directly use the AI ​​big model based on the information fusion results and the information of multiple products in the product library to match and give search results. In this way, since the AI ​​big model can process multimodal information such as text, pictures, and videos, it may improve the accuracy of search results. However, since the AI ​​big model may take a long time to perform the above matching process, users may need to wait for a long time to obtain search results. Alternatively, in actual applications, some predictive work can be done in advance on the user's possible search needs, and the AI ​​big model can match the search needs with the products in an offline manner, and store the matching results, so that after collecting and analyzing the user's actual search demand information, the search results can be returned by querying the stored information, and so on.

[0069] In summary, in an embodiment of the present application, an input box can be provided in the first area (i.e., the LUI part) to facilitate multiple rounds of interaction between the user and the AI ​​assistant. When the user has multiple rounds of natural language interaction, the system can model the multiple rounds of context through the rewriting query module to ensure that the current user's search needs are comprehensive and complete, thereby collecting fine-grained search needs. Of course, the system can also correct problems such as inaccurate user expressions and spelling errors.

[0070] In addition, since the control panel in the second area (i.e., the GUI part) can provide a list of attributes and attribute values ​​that are highly relevant to the current needs, users can quickly express their needs by clicking. Once the user selects the content, the system will collect the information for search matching. After obtaining the search results, the system will update a new batch of search demand suggestions for the user to further select.

[0071] Finally, the demand fusion part can rewrite the unstructured expressions collected by the LUI and then merge them with the structured requirements collected by the GUI. It can normalize the information expressed in natural language and combine it with the user's selection content in the console to achieve a more comprehensive understanding of the requirements and achieve better matching results. After the demand fusion module integrates the requirements of the LUI and GUI, it can generate search query statements and filter items and send them to the search engine query results.

[0072] After obtaining the search results, the system can comprehensively consider the demand understanding results and search results, and use the AI ​​big model to analyze the user's current demand satisfaction, potential search intent and other dimensions to update the content. This includes natural language feedback and prompts as well as the console candidate search demand content, so as to prompt the user with matching conditions and further guide the user to the next search behavior.

[0073] Through the embodiments of the present application, since a first area (LUI, natural language user interface) and a second area (GUI, graphical user interface) can be provided in the search interaction interface, the user can input natural language sentences through the LUI and select attributes and attribute values ​​through the GUI during the search process. Therefore, the flexibility of the LUI can be used to give users the opportunity to express their search needs in a fine-grained and delicate manner, and to clarify, refine, and clarify their own search needs during multiple rounds of dialogue; at the same time, the GUI is used to provide visual search demand options and provide candidates for user needs, so that users can quickly and efficiently adjust and refine search conditions. This design is conducive to improving user experience, reducing the possibility of input errors, and enhancing matching efficiency. Afterwards, the information collected from the LUI and GUI can be integrated and the corresponding search results can be given.

[0074] Compared with traditional product search, the solution provided by the embodiment of the present application can at least include the following effects:

[0075] (1) Solving expression limitations: Traditional search boxes limit the input length and are difficult to fully express fine-grained requirements. In the solution of the embodiment of the present application, the LUI allows users to express freely in unstructured text, and at the same time combines the visualization options of the GUI to facilitate users to quickly refine and adjust their requirements and improve matching efficiency.

[0076] (2) Enhanced expression accuracy: Users may not be able to express their intentions clearly due to insufficient professional terminology or typos. The solution of the embodiment of the present application analyzes user input through multiple rounds of interactions and refers to historical information, combined with the structured options of the GUI, to ensure the accuracy of demand collection and reduce the risk of mismatching.

[0077] (3) Reducing user burden: Traditional solutions require users to actively think about the next step of the search, which increases the psychological burden. The solution of the embodiment of the present application combines the multi-round dialogue of the LUI and the quick options of the GUI, allowing users to gradually clarify their needs, effectively advancing the search process and reducing the burden.

[0078] (4) Demand fusion: The demand fusion part can integrate the information collected from the LUI and the GUI, and conduct a comprehensive analysis of the structured information collected by the GUI and the unstructured input collected by the LUI to ensure that the generated search query is comprehensive and accurate, covering the user's current and historical needs.

[0079] (5) Under the preferred implementation mode, real-time optimization and feedback can be performed: after executing the search, the relationship between the "user demand understanding results" and the "search matching results" can be analyzed, the prompt content can be automatically updated, and effective next search prompts can be provided, including search suggestions in the GUI (attribute and attribute value options) and / or natural language feedback in the LUI (that is, the reply content of the AI ​​large model). This feedback mechanism can help users quickly adjust their search strategies, effectively guide subsequent operations, improve the overall efficiency of the search, and continuously optimize the user experience.

[0080] Embodiment 2

[0081] This second embodiment provides a product search method from the perspective of the server. Figure 5 , the method may include:

[0082] S501: receiving a product search request submitted by a client side, wherein the product search request is generated after the client collects user expression information on search requirements through a product search interactive interface, wherein the interface includes a first area and a second area, wherein the first area is used to provide an AI interaction component to receive a user expressing search requirements by inputting natural language sentences, and to receive a sentence input by the user inspired by the reply content of the AI ​​big model to clarify the search requirements through multiple rounds of interaction with the AI ​​big model; the second area is used to provide a plurality of attribute options for clarifying the search requirements and a plurality of attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search requirement information expressed by the user by selecting the attributes and attribute values;

[0083] S502: fusing the input sentences collected through the first area and the attributes and attribute value selection results collected through the second area;

[0084] S503: Generate product search results based on the information fusion results, and return them to the client for display.

[0085] For the contents not described in detail in the second embodiment, please refer to the records in the first embodiment and other parts of this specification, which will not be repeated here.

[0086] It should be noted that the embodiments of the present application may involve the use of user data. In actual applications, user-specific personal data can be used in the scheme described herein within the scope permitted by applicable laws and regulations, subject to the requirements of applicable laws and regulations of the country where the user is located (for example, with the user's explicit consent, effective notification to the user, etc.).

[0087] Corresponding to the first embodiment, the embodiment of the present application further provides a commodity search device, which may include:

[0088] An interface providing unit is used to provide a commodity search interactive interface, wherein the interface includes a first area and a second area, wherein the first area is used to provide an artificial intelligence (AI) interactive component to receive a user's expression of search needs by inputting natural language sentences, and to receive, through multiple rounds of interaction with the AI ​​big model, sentences input by the user inspired by the reply content of the AI ​​big model to clarify the search needs; the second area is used to provide a plurality of attribute options for clarifying the search needs and a plurality of attribute value options corresponding to the attribute options according to the interaction situation in the first area, to receive the search demand information expressed by the user by selecting the attributes and attribute values;

[0089] A search result providing unit is used to provide product search results based on the information collected through the first area and the second area. The product search results are generated based on the information fusion result after the input sentence collected through the first area and the attribute and attribute value selection results collected through the second area are fused.

[0090] In a specific implementation, the device may further include:

[0091] A rewriting unit is used to rewrite the attribute and attribute value selection results into sentences expressed in natural language after collecting the attribute and attribute value selection results through the second area, and display them in the dialogue area of ​​the first area.

[0092] Among them, when performing the information fusion, the multiple rounds of input statements collected through the first area can be rewritten into structured search query statements suitable for the search engine, and fused with the attribute and attribute value selection results collected through the second area to generate a complete structured search query statement, so as to generate corresponding search results based on the structured search query statement.

[0093] Among them, when generating the complete structured search query statement, the AI ​​big model can be used to perform natural language understanding on the multiple rounds of input statements and the attribute and attribute value selection results, so as to rewrite the multiple rounds of input statements based on understanding the user's search needs and generate the complete structured search query statement.

[0094] Additionally, it may include:

[0095] The updating unit is used to analyze the matching between the search results and the understanding results of the user's search needs through the AI ​​big model after obtaining the search results, and update the feedback content of the first area and / or the attribute options and attribute value options displayed in the second area.

[0096] The user includes merchant buyer users, and the merchant buyer users include: buyer users who purchase or customize commodities in bulk through the current commodity information service system for resale or production purposes;

[0097] The search requirements include: the requirement for group purchase of multiple different minimum stock keeping units (SKUs) of the same product; wherein the SKUs are expressed by combining attribute values ​​in multiple dimensions;

[0098] In the process of collecting search requirements, multiple selections of different attribute values ​​under the same attribute are supported.

[0099] In a preferred implementation, when generating the complete structured search query statement, a plurality of different selected attribute values ​​under the same attribute are set to a logical OR relationship, so that when providing corresponding product search results, products that can meet the user's purchasing needs in terms of corresponding SKU supply are screened from the product library based on the logical OR relationship, and are displayed preferentially in the product search results.

[0100] Corresponding to the second embodiment, the embodiment of the present application further provides a commodity search device, which may include:

[0101] A request receiving unit, used to receive a commodity search request submitted by a client side, wherein the commodity search request is generated after the client collects the user's expression information on the search demand through the commodity search interactive interface, wherein the interface includes a first area and a second area, wherein the first area is used to provide an AI interaction component to receive the user's expression of the search demand by inputting a natural language sentence, and to receive the sentence input by the user under the inspiration of the reply content of the AI ​​big model to clarify the search demand through multiple rounds of interaction with the AI ​​big model; the second area is used to provide a plurality of attribute options for clarifying the search demand and a plurality of attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search demand information expressed by the user by selecting the attributes and attribute values;

[0102] An information fusion processing unit, used for fusing the input sentences collected through the first area and the attributes and attribute value selection results collected through the second area;

[0103] The search result generation unit is used to generate product search results based on the information fusion results and return them to the client for display.

[0104] In addition, an embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the steps of any one of the methods in the aforementioned method embodiments are implemented.

[0105] And an electronic device, comprising:

[0106] one or more processors; and

[0107] A memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of the method described in any one of the aforementioned method embodiments.

[0108] A computer program product includes a computer program / computer executable instructions, which, when executed by a processor in an electronic device, implement the steps of the method described in the aforementioned method embodiment.

[0109] in, Figure 6 The architecture of the electronic device is shown as an example, which may include a processor 610, a video display adapter 611, a disk drive 612, an input / output interface 613, a network interface 614, and a memory 620. The processor 610, the video display adapter 611, the disk drive 612, the input / output interface 613, the network interface 614, and the memory 620 may be communicatively connected via a communication bus 630.

[0110] Among them, the processor 610 can be implemented by a general-purpose CPU (Central Processing Unit, processor), a microprocessor, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, etc., to execute relevant programs to implement the technical solution provided in this application.

[0111] The memory 620 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 620 can store an operating system 621 for controlling the operation of the electronic device 600, and a basic input and output system (BIOS) for controlling the low-level operation of the electronic device 600. In addition, a web browser 623, a data storage management system 624, and a commodity search processing system 625, etc. can also be stored. The above-mentioned commodity search processing system 625 can be an application program that specifically implements the operations of the aforementioned steps in the embodiment of the present application. In short, when the technical solution provided by the present application is implemented by software or firmware, the relevant program code is stored in the memory 620 and is called and executed by the processor 610.

[0112] The input / output interface 613 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0113] The network interface 614 is used to connect to a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0114] The bus 630 comprises a pathway for transmitting information between the various components of the device (eg, the processor 610, the video display adapter 611, the disk drive 612, the input / output interface 613, the network interface 614, and the memory 620).

[0115] It should be noted that, although the above device only shows a processor 610, a video display adapter 611, a disk drive 612, an input / output interface 613, a network interface 614, a memory 620, a bus 630, etc., in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include components necessary for implementing the solution of the present application, and does not necessarily include all the components shown in the figure.

[0116] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly contributed to the prior art in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0117] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0118] The above is a detailed introduction to the commodity search method and electronic device provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and its core idea of ​​the present application. At the same time, for those skilled in the art, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present application.

Claims

1. A commodity search method, characterized in that: include: A commodity search interactive interface is provided, the interface including a first area and a second area, the first area is used to provide an artificial intelligence (AI) interactive component to receive search requirements expressed by a user by inputting natural language sentences, and to receive sentences input by the user inspired by the reply content of the AI ​​big model to clarify the search requirements through multiple rounds of interaction with the AI ​​big model; the second area is used to provide multiple attribute options for clarifying the search requirements and multiple attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search requirement information expressed by the user by selecting the attributes and attribute values; Product search results are provided based on the information collected through the first area and the second area. The product search results are generated based on the information fusion result after the input sentence collected through the first area and the attribute and attribute value selection results collected through the second area are fused.

2. The method according to claim 1, characterized in that Also includes: After the attributes and attribute value selection results are collected through the second area, the attributes and attribute value selection results are rewritten into sentences expressed in natural language and displayed in the dialogue area of ​​the first area.

3. The method according to claim 1, characterized in that When the information fusion is performed, it includes: The multiple rounds of input statements collected through the first area are rewritten into structured search query statements suitable for search engines, and are merged with the attribute and attribute value selection results collected through the second area to generate a complete structured search query statement, so as to generate corresponding search results according to the structured search query statement.

4. The method according to claim 3, characterized in that When generating the complete structured search query statement, the AI ​​big model performs natural language understanding on the multiple rounds of input statements and the attribute and attribute value selection results, so that based on understanding the user's search needs, the multiple rounds of input statements are rewritten and the complete structured search query statement is generated.

5. The method according to claim 4, characterized in that Also includes: After obtaining the search results, the AI ​​big model is used to analyze the match between the search results and the results of understanding the user's search needs, and the feedback content in the first area and / or the attribute options and attribute value options displayed in the second area are updated.

6. The method according to claim 1, characterized in that The users include merchant buyer users, and the merchant buyer users include: buyer users who purchase or customize commodities in bulk through the current commodity information service system for resale or production purposes; The search requirements include: the requirement for group purchase of multiple different minimum stock keeping units (SKUs) of the same product; wherein the SKUs are expressed by combining attribute values ​​in multiple dimensions; In the process of collecting search requirements, multiple selections of different attribute values ​​under the same attribute are supported.

7. The method according to claim 6, characterized in that When generating the complete structured search query statement, multiple different attribute values ​​selected under the same attribute are set to a logical OR relationship, so that when providing corresponding product search results, products that can meet the user's purchasing needs in terms of corresponding SKU supply are screened from the product library based on the logical OR relationship, and are displayed preferentially in the product search results.

8. A commodity search method, characterized in that: include: Receive a product search request submitted by a client side, wherein the product search request is generated after the client collects user expression information on search needs through a product search interactive interface, wherein the interface includes a first area and a second area, wherein the first area is used to provide an AI interaction component to receive the search needs expressed by the user by inputting natural language sentences, and receive, through multiple rounds of interaction with the AI ​​big model, sentences input by the user inspired by the reply content of the AI ​​big model to clarify the search needs; the second area is used to provide multiple attribute options for clarifying the search needs and multiple attribute value options corresponding to the attribute options according to the interaction situation in the first area, so as to receive the search demand information expressed by the user by selecting the attributes and attribute values; Performing information fusion on the input sentences collected through the first area and the attributes and attribute value selection results collected through the second area; Generate product search results based on the information fusion results and return them to the client for display.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 8 are implemented.

10. An electronic device, characterized in that: include: one or more processors; as well as A memory associated with the one or more processors, the memory being used to store program instructions, wherein the program instructions, when read and executed by the one or more processors, execute the steps of the method described in any one of claims 1 to 8.

11. A computer program product comprising a computer program / computer executable instructions, characterized in that: When the computer program / computer executable instructions are executed by a processor in an electronic device, the steps of the method according to any one of claims 1 to 8 are implemented.

Images