Commodity search processing method and electronic device

Abstract

The present application discloses a product search processing method and electronic device, the method comprising: obtaining operational behavior information related to positive and / or negative feedback generated by a user during the process of providing product search results based on current search keywords; determining a target tag word from the tag words of the products in the product search results based on the operational behavior information; and recommending the target tag word to the user so as to guide the user to provide a more accurate search keyword through the target tag word. Through the present application embodiment, the user gains a sense of participation while also making the search results more in line with user needs.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of commodity search, in particular to a commodity search processing method and an electronic device. BACKGROUND

[0002] In a commodity information service system, a commodity search function can be provided for a user, so that the user can search for a desired commodity by inputting a search keyword. After receiving a search request, a search engine can determine matching commodities according to the keyword, perform algorithm analysis according to historical consumption habits, browsing records and other data of the user, and return the matching commodities to the user after sorting.

[0003] However, the keywords input by the user in the search process are usually large category words or broad words, such as “sweaters” and “coats”. The accuracy of the results obtained by algorithm processing based on such keywords is low, and the search efficiency is low.

[0004] Some navigation-type help information exists in the prior art. For example, after the user inputs a search word such as “sweaters”, a navigation bar can be provided at the top of a search result page according to the sub-categories under the category of sweaters. The navigation bar provides a plurality of navigation words, such as navigation words of sub-categories from the dimensions of material, style and style. The user can select a navigation word from the navigation bar to add it to the keyword, so as to perform accurate search for the desired commodity in a smaller range corresponding to the navigation word.

[0005] The above-mentioned method can help the user to improve the search efficiency to some extent. However, the number of navigation words is large, and even the complete navigation words cannot be displayed on one screen. The user needs to horizontally slide the navigation bar to browse all the navigation words. At this time, the user may face the problem of not knowing how to select from the numerous navigation words, or after seeing the first few navigation words, if the user does not find the desired word, the user may directly ignore the navigation words, so that the navigation words are useless.

[0006] Therefore, how to provide more practical help for the user in the commodity search process has become a technical problem to be solved by those skilled in the art. SUMMARY

[0007] The present application provides a commodity search processing method and an electronic device, which can make the user feel involved and make the search results more meet the user's needs.

[0008] The present application provides the following solutions:

[0009] A commodity search processing method, comprising:

[0010] In the process of providing the commodity search result based on the current search keyword, operation behavior information related to positive feedback and / or negative feedback generated by the user is acquired;

[0011] According to the operation behavior information, a target label word is determined from label words possessed by commodities in the commodity search result;

[0012] The target label word is recommended to the user, so as to guide the user to provide a more accurate search keyword through the target label word.

[0013] The operation behavior information related to positive feedback and / or negative feedback generated by the user is acquired, including:

[0014] Operation behavior information related to positive feedback and / or negative feedback generated by the user on the commodity search result is acquired.

[0015] The operation behavior information related to positive feedback and / or negative feedback generated by the user on the commodity search result is acquired, including:

[0016] If the user performs a view detail information operation on a commodity in the commodity search result, the commodity is determined as a positive commodity obtaining positive feedback of the user;

[0017] The target label word is determined from label words possessed by commodities in the commodity search result according to the operation behavior information, including:

[0018] When returning to the commodity search result page, a target label word is selected from label words possessed by the positive commodity, and the target label word is used as the recommended target label word.

[0019] The operation behavior information related to positive feedback and / or negative feedback generated by the user on the commodity search result is acquired, including:

[0020] If the user continuously slides multiple screens on the search result page without a click operation, commodities displayed through the sliding operation are determined as multiple negative commodities obtaining negative feedback of the user;

[0021] The target label word is determined from label words possessed by commodities in the commodity search result according to the operation behavior information, including:

[0022] A label word having a commonality possessed by the multiple negative commodities is determined as a negative label word;

[0023] After the negative label word is removed from multiple label words possessed by commodities in the commodity search result, the recommended target label word is selected from the remaining label words.

[0024] The method further includes:

[0025] After obtaining the new positive feedback and / or negative feedback related operation behavior information, the recommended target label word is updated.

[0026] Further comprising:

[0027] After receiving the target operation performed by the user on the recommended target label word, the target label word is combined with the current search keyword to form a new search keyword;

[0028] The product search result is refreshed according to the new search keyword.

[0029] The recommended target label word is displayed within a threshold time period.

[0030] The method further comprises:

[0031] If no target operation is performed by the user on the recommended target label word within the threshold time period, the user is provided with a recommended label word according to the new operation behavior of the user and / or the operation behavior generated in the current search process.

[0032] Further comprising:

[0033] The label word that has been recommended but on which no target operation is performed by the user is determined as a negative label word.

[0034] When the user is provided with a recommended label word again, the negative label word is removed from the plurality of label words possessed by the product search result, and the recommended label word is selected from the remaining label words.

[0035] The recommended target label word is displayed in the form of floating subtitles within a threshold time period.

[0036] A single recommended label word is displayed within the same threshold time period.

[0037] The target label word comprises a label word obtained by analyzing and mining user-generated content related to product description, and adding a label to a related product in a product information service system according to product information associated with the user-generated content, the user-generated content including text, short video and / or live content.

[0038] The target label word comprises a label word obtained by obtaining a description word describing the characteristics of the product in use from the user-generated content related to the product and clustering the description word.

[0039] And / or,

[0040] From the user-generated content related to the goods, determine relevant topic discussion content, and mine label words belonging to discussion hotspots or trends from the topic discussion content.

[0041] Among them, the label words obtained by analyzing and mining the user-generated content also include exclusive label words corresponding to a single product, and the exclusive label words include label words related to celebrity same products or the number of people interested;

[0042] The method further comprises:

[0043] When displaying the product search results, display the exclusive label words related to celebrity same products or the number of people interested corresponding to the product search results to assist users in making selection decisions from the product search results.

[0044] Among them, it also includes:

[0045] According to the user historical behavior data corresponding to the product, the label words possessed by the product, and the user's group label information, determine the preference information of each group for the label words;

[0046] The target label word is determined from the label words possessed by the goods in the product search results according to the operation behavior information, comprising:

[0047] Determine the target group to which the current search user belongs;

[0048] According to the label words possessed by the goods in the product search results, the label words preferred by the target group, and the operation behavior information, determine the target label word.

[0049] A product search processing method, comprising:

[0050] Based on the current search keyword, provide product search results;

[0051] In the process of providing the search results, provide recommended label words, wherein the label words are obtained by analyzing and mining user-generated content related to product information;

[0052] After the recommended label words are selected, combine the current search keyword and the recommended label words into new search keywords, and after re-updating the search, provide new product search results.

[0053] Among them, the label words are associated with product categories, and a plurality of label words are included under the same product category;

[0054] The method further comprises:

[0055] determine a target commodity category according to the current search keyword;

[0056] determine a target label word according to a label word associated with the target commodity category.

[0057] wherein the method further comprises:

[0058] determine preference information of each target user group for a label word associated with each commodity category;

[0059] determine a target label word according to a label word associated with the target commodity category, comprising:

[0060] determine a target user group to which a searcher belongs;

[0061] determine the target label word according to a label word preferred by the target user group under the target commodity category.

[0062] wherein the recommended label word is displayed within a threshold time period;

[0063] the method further comprises:

[0064] if no selection operation of the recommended label word is received from the user within the threshold time period, recommend a new label word to the user.

[0065] A processing device for commodity search, comprising:

[0066] an operation behavior information acquisition unit configured to acquire operation behavior information related to positive feedback and / or negative feedback generated by a user in a process of providing a commodity search result based on a current search keyword;

[0067] a label word determination unit configured to determine a target label word from label words possessed by commodities in the commodity search result according to the operation behavior information;

[0068] a label word recommendation unit configured to recommend the target label word to the user so as to guide the user to provide a more accurate search keyword through the target label word.

[0069] A processing device for commodity search, comprising:

[0070] a search result providing unit configured to provide a commodity search result based on a current search keyword;

[0071] a label word recommendation unit configured to provide a recommended label word in a process of providing the search result, wherein the label word is obtained by analyzing and mining user-generated content related to commodity information;

[0072] A re-searching unit is configured to combine the current search keyword and the recommended label word as a new search keyword after the recommended label word is selected, and to re-update the search to provide a new product search result.

[0073] A computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method of any preceding method.

[0074] An electronic device comprising:

[0075] One or more processors; and

[0076] A memory associated with the one or more processors, the memory configured to store program instructions that, when executed by the one or more processors, perform the steps of the method of any preceding method.

[0077] According to the embodiments provided in the present application, the following technical effects are disclosed:

[0078] According to the embodiments provided in the present application, in the process of providing a product search result based on a current search keyword, operation behavior information related to positive feedback and / or negative feedback generated by a user can be collected, for example, positive feedback behavior such as a user clicking to view the detail information of a product, or negative feedback behavior such as continuous sliding without clicking, and the like. Then, according to the operation behavior information, a target label word can be selected from the label words possessed by the products in the product search result, and recommended to the user, so as to guide the user to provide a more accurate search keyword through the target label word. That is, in the embodiments of the present application, some optional label words can also be recommended to the user to guide the user to perform a more accurate search, but the specific label words to be displayed are determined according to the positive feedback and / or negative feedback reflected by the operation behavior of the user in the search process, so as to form a recommended label in an interactive form, rather than simply listing a plurality of label words in a navigation bar. In other words, in the embodiments of the present application, the specific recommended label word is generated along with the operation behavior of the user, and through the positive feedback or negative feedback behavior continuously generated by the user, a label word that is more likely to meet the demand of the user can be gradually recommended. Moreover, the user is allowed to select whether to use a specific label word to perform a more accurate search, rather than automatically selecting in the background of the program, so as to give the user the initiative to make a selection, so that the user obtains a sense of participation, and the search result is also more likely to meet the demand of the user.

[0079] Among them, regarding the label words of specific recommendations, the label words can be pre-analyzed from the user production content (which can include text and image content, short video, live content, etc.) and mined out, to be used to enrich the label words generated based on the attributes of the commodity. The label words mined out in this way are closer to the user's experience or experience summary of the commodity in the actual life scene, and therefore can have the characteristics of being more practical, more comprehensive, and more reflecting the current popular trend, and thus can help the user better express the search demand.

[0080] Of course, implementing any product of the present application does not necessarily require all the advantages described above. BRIEF DESCRIPTION OF DRAWINGS

[0081] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0082] Figure 1 is a schematic diagram of the system architecture provided by the embodiments of the present application;

[0083] Figure 2 is a flowchart of the first method provided by the embodiments of the present application;

[0084] Figure 3 is a schematic diagram of the interface provided by the embodiments of the present application;

[0085] Figure 4 is a flowchart of the second method provided by the embodiments of the present application;

[0086] Figure 5 is a schematic diagram of the first device provided by the embodiments of the present application;

[0087] Figure 6 is a schematic diagram of the second device provided by the embodiments of the present application;

[0088] Figure 7 is a schematic diagram of the electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0089] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art belong to the scope of protection of the present application.

[0090] The inventors of the present application find in the process of implementing the present application that the user in the commodity information service system usually has certain appeal when initiating the commodity search, but may not be sure how to accurately express the appeal, and thus inputs a relatively broad keyword for search. However, this search method is often difficult to search for the real desired commodity.

[0091] Therefore, in the embodiments of the present application, an implementation scheme of "companion search" is provided, in which operation behavior information of the user can be collected in the process of providing the commodity search result based on the current search keyword. This operation behavior information can be regarded as positive feedback and / or negative feedback of the user. For example, if the user clicks a certain commodity in the search result for detail viewing, the behavior belongs to positive feedback, proving that the commodity can be more in line with the needs of the user. If the user keeps sliding the search result page without clicking action, it belongs to negative feedback, proving that the user can not be interested in the search result that has been exposed, and the like. In addition, a specific commodity can usually have some label words, and different commodities with the same label words can have commonality in a certain aspect. Therefore, the target label word can be determined from the label words of the commodities in the search result based on the positive feedback and / or negative feedback information, and is displayed to the user. If the user is interested in the target label word, it can be added to the search box by clicking or the like, so as to be used as a supplement to the current keyword, thereby narrowing the search range and facilitating the acquisition of more suitable search result.

[0092] For example, assuming that the user clicks a certain commodity in the search result, the target label word can be determined from the label words of the commodity and recommended to the user when the feedback is made to the search result page. If the user keeps sliding the search result page (for example, has slid two screens) and does not produce clicking or the like, the target label word can also be recommended to the user at this time. In this case, after the common label words are determined from the label words of the commodities that have been exposed and are removed from the multiple label words of the commodities in the search result, a certain target label word is determined and recommended to the user. This recommended target label word can be displayed only for a relatively short period of time, for example, a few seconds, and the like. If the user does not perform clicking operation on the target label word within the time period, other label words can also be recommended.

[0093] In summary, in the embodiments of the present application, the specific target label word can be dynamically recommended according to the positive feedback and / or negative feedback behaviors generated by the user in the process of browsing the search results. That is, the recommended label word is related to the operation behaviors generated by the user in the current search process, and the recommended label word can be recommended and updated multiple times as the specific positive and / or negative feedback operation behaviors are continuously generated. The user can click on the recommended label word, so that the search keyword is more accurate, and the search result will be more and more accurate. Therefore, the user can obtain the experience that the search engine is always helping him to input more accurate keywords, and therefore, it can be called "accompanying search".

[0094] It should be noted here that in some prior art, after the user performs some behaviors in the process of browsing the search results, the search engine can also consider this behavior information for updating the search results. However, the processing manner in the prior art belongs to automatically updating the search results after the program background analyzes the user behavior data, but the user cannot feel the process of the algorithm, and therefore, a sense of distrust of the algorithm can be generated. In the embodiments of the present application, the target label word is determined according to the positive feedback and / or negative feedback information of the user, and is explicitly recommended to the user for selection of whether to use the target label word, so that the selection right is returned to the user, and the final result is more in line with the real needs of the user.

[0095] Among them, the label word of the specific commodity can have multiple acquisition manners. For example, it can be determined according to the attribute classification information of the commodity itself, for example, the label word can be added to the commodity in the style, category and other dimensions. However, in the embodiments of the present application, considering that the corresponding appeal of the user in the process of searching for commodities can be a relatively practical appeal, especially some relatively personalized appeals, it is often difficult to express the appeal of the user through the attribute classification of the commodity itself, which leads to the difficulty in searching for the desired commodity. For example, the user wants to buy a sweater suitable for attending a certain occasion. At this time, if "sweater" is used as a keyword to initiate a search, although the prior art can provide some fine categories under the "sweater" category for the user to select, for example, from the perspective of style, it can provide multiple selectable fine categories such as round neck sweater, V-neck sweater, and fake two-piece sweater. However, the user can have difficulty in determining which specific style and style are suitable for attending the occasion he needs to attend, and therefore, the navigation information provided from the attribute of the commodity itself cannot effectively help the user express the above appeal.

[0096] On the other hand, some content publishing systems (provide a corresponding platform for the publication of content produced by users (may include the production of content by various users, including ordinary users, professional users, etc.), various users can produce content in the content publishing system, including short video, live content, etc.) often include some user-produced content related to goods. For example, some users publish short video content on the theme of "daily dressing", etc. Since such user-produced content is usually summarized from some feelings, experiences or experiences in the process of actually using specific goods, the occasions suitable for specific goods may have more specific introductions. The style introduction of the goods may not be limited to some common ways, but follow the trend to produce some more innovative expression ways. In addition, the publishers of these contents may also include some famous people, including some "stars", "big Vs", etc. If these famous people recommend a certain product, it may be more "authoritative" information for other users, and users may be more willing to buy the same product, etc.

[0097] In summary, such user-produced content may be more useful to other users. Therefore, in actual application, there may be a part of users, especially young users, who have the following phenomenon: before initiating a search through the commodity information service system, they first browse the content publishing system to see which goods are recommended by other users or stars, big Vs, etc. or which goods are popular or trendy, etc. According to their own needs, they obtain "knowledge" about clothing matching, fashion trends, etc. and then go to the commodity information service system to search for goods more purposefully. However, this way will obviously make the path for users to obtain the final goods they want to purchase longer, and if the information in the commodity information service system cannot be better matched and integrated with the knowledge obtained in the content publishing system, it may also be difficult to search for the desired goods. For example, suppose a user sees a certain style of sweater in the content publishing system, but the user-produced content describes the style as being quite trendy. The commodity information service system may have a similar sweater, but it does not include the relevant style description words. At this time, if the user directly uses the aforementioned quite trendy style description words to initiate a search in the commodity information service system, it may be difficult to match the search results that meet the conditions.

[0098] Based on the above, in the embodiments of the present application, the user-generated content related to the goods in at least one content publishing system can be analyzed in advance, and some label words more inclined to user emotional expression can be mined from the user-generated content. Since the user-generated content is also related to the goods, the label words are also related to some target goods. Accordingly, the mined label words can be used to add corresponding labels to the goods matching the target goods in the goods information service system. In this way, the goods in the goods information service system can have some label words related to the attributes of the goods, and some label words mined from the user-generated content in the content publishing system, which are not related to the attributes of the goods. These more abundant label words can be used to provide auxiliary information for user search.

[0099] In the process of label word mining, the goods categories defined in the category system of the goods information service system can be mined. That is, if the goods information service system includes multiple goods categories, the user-generated content can be collected and label word mining can be performed for different goods categories. In specific implementation, the specific label words can correspond to different label dimensions, such as scene and style. The label dimensions in which label word mining is required under different goods categories can be configured by relevant personnel such as operators in advance. Then, relevant algorithms can be used to perform voice recognition, semantic analysis, and other processing on the specific user-generated content to mine relevant label words.

[0100] The specific label words can be mainly divided into two types. One type of label word corresponds to a type of goods, that is, the user adds a classification label to the goods, such as label words in the scene dimension, label words in the style dimension, and the like. One label word usually corresponds to multiple goods. The other type of label word corresponds to a single good, that is, a label can be added to a specific good, such as a good recommended by a celebrity, or the number of people interested in a good, and the like.

[0101] In providing the search auxiliary information for the user by using the above various different label words, there can be various manners. For example, in one manner, the label words used for classifying the goods can be used to help the user to gradually narrow the search range and improve the accuracy of the search result. At this time, the operation behavior information related to the positive feedback and / or the negative feedback generated by the user can be acquired in the process that the user browses the search result, and the target label word can be determined from the label words possessed by the goods in the search result according to the operation behavior information, and recommended to the user. If the user clicks to select the target label word, the target label word can be added to the search input control as a supplement to the search keyword, and the search result is refreshed to realize the filtering or reordering of the search result. Since the label words provided by the embodiments of the present application can be obtained by mining from the user production content, they can more express the characteristics of the goods from the user's perceptual angle, so that the recommended label words are more practical and more conducive to expressing the practical search appeal of the searcher user, thereby more effectively helping to improve the accuracy of the search result.

[0102] In addition, for the label words corresponding to a specific good, the label words can be directly displayed in the search result. For example, if a good is recommended by a star, or is interested by many people, the label words such as "recommended by a star" and "a certain number of people have planted grass" can be provided at the resource position of the good in the search result page. In this way, the label words can help the searcher to make a decision on the selection of the good.

[0103] In this way, since some label words can be mined from the user production content of the content publishing system to be used to add labels to the goods matched in the good information system, the user production content in the content publishing system and the goods in the good information system are better integrated. In the process that the user searches for goods, the label words can be used to provide search auxiliary information for the user to help improve the accuracy of the search result, or to help make a decision on the selection of the good, etc. Therefore, the problem of long operation path caused by switching between various different systems is avoided.

[0104] From the system architecture point of view, as shown in Figure 1 the embodiments of the present application can provide the search result through the client of the good information service system, and collect the operation behavior information related to the positive feedback and / or the negative feedback generated by the user. The operation behavior information can be submitted to the server, and the target label word can be determined from the label words possessed by the goods in the search result according to the operation behavior information by the server. Then, the client can recommend the target label word to the user. If the user accepts the recommendation, the target label word can be added to the keyword to reinitiate the search.

[0105] In order to obtain more rich label words about specific goods, or label words that describe the goods from a more practical perspective, user-generated content related to the goods can also be obtained from multiple content publishing systems (which can be third-party content publishing systems, or content publishing systems associated internally with the goods information service system, etc.) including text and images, short videos, live broadcasts, etc. Then the label mining algorithm in the goods information service system can be used to analyze and mine label words from the user-generated content. For example, multiple classification label words can be mined from the dimensions of scene and style, or multiple label words for a single good can be mined from the dimensions of celebrity recommendation and multiple people "grass-roots", etc. After the specific label words are mined, since the label words are also associated with the goods in the user-generated content, the goods information library of the goods information service system can be searched to find goods related to the goods (same or similar, etc.), and the corresponding label can be added to the matched goods using the label. In this way, when receiving a user search request through a search engine, the label can be used to provide auxiliary information to the searcher user. For example, the classification label words can be used to help the user narrow down the search range and improve the accuracy of the search results. Or, according to the label words corresponding to a single good, the user can make a decision on the selection of goods, etc.

[0106] It should be noted that the goods information service system described in the embodiments of the present application can be a traditional comprehensive goods information service system, that is, the goods categories covered can be very comprehensive, and the consumer users can also include a variety of different user groups; or it can also be a goods information service system for some vertical fields, for example, it can focus on providing services for some fashion clothing and other related categories of goods, and the consumer users it faces are mainly young groups, etc. In actual application, the scheme provided by the embodiments of the present application can be more suitable for implementation in the latter scenario, because the young group of users is more likely to have personalized goods search needs and more likely to accept information mined from user-generated content. In addition, when providing search services for users, the goods information service system for vertical fields pays more attention to the accuracy of the search results, so there is a higher demand for classification of goods from multiple different dimensions.

[0107] The specific implementation scheme provided by the embodiments of the present application will be described in detail below.

[0108] Embodiment one

[0109] Specifically, the first embodiment of the present application first provides a goods search processing method, which is described in detail with reference to Figure 2 The method can include:

[0110] S201: In the process of providing the commodity search result based on the current search keyword, operation behavior information related to positive feedback and / or negative feedback generated by the user is acquired.

[0111] Specifically, the user in the commodity information service system can initiate a commodity search request by inputting a keyword in a relevant page provided by the client. The keyword input by the user is usually some broad, general, and large category words, such as “sweater” and “down jacket”.

[0112] After receiving the search request, the search result corresponding to the search keyword can be acquired first, which can include information of a plurality of commodities matching the search keyword. It should be noted that in an optional manner, the search result can be determined and sorted according to the matching degree of the commodity and the keyword in the process of providing the search result according to the keyword input by the user in the embodiment of the application, without considering the style and season preferred by the user too much. As for such factors, the user can select them in the subsequent interaction process with the user.

[0113] In the embodiment of the application, in the process of showing the search result to the user, operation behavior information related to positive feedback and / or negative feedback generated by the user can be acquired. Specifically, it can include operation behavior information related to positive feedback and / or negative feedback generated by the user to the commodity search result, for example, whether the user clicks a certain commodity in the search result to view the details, or whether the user slides multiple screens without clicking, etc. According to the specific operation behavior type, it can be determined whether it is positive feedback or negative feedback. For example, for the case of clicking a certain commodity to view the details, it belongs to positive feedback, and for the case of continuously sliding multiple screens without clicking, it belongs to negative feedback, etc.

[0114] S202: According to the operation behavior information, a target label word is determined from label words possessed by commodities in the commodity search result.

[0115] In the process of providing the search result, the user can be provided with help by providing the user with recommended label words, and the label words can be determined according to the collected operation behaviors of the user in the current search process. For example, if the user performs a view detail information operation on a product in the product search result, the product can be determined as a positive product that obtains positive feedback from the user. At this time, when returning to the product search result page, a target label word can be selected from the label words possessed by the positive product as the recommended target label word. For example, assuming that the user clicks a product to view the detail page, it proves that the user can be interested in the style of the product, and therefore a target label word can be determined from the label words possessed by the product for recommendation to the user. According to further feedback of the user, it is determined whether the target label word meets the needs of the user. If so, the search range is narrowed based on the target label word, otherwise, other label words can be recommended to the user, and so on.

[0116] Alternatively, if the user continuously slides multiple screens on the search result page without a click operation, the products displayed by the sliding are determined as multiple negative products that obtain negative feedback from the user, and the label words possessed by the multiple negative products that have commonality can be determined as negative label words. Then, the negative label words are removed from the multiple label words possessed by the products in the product search result, and the recommended target label word is selected from the remaining label words. That is, assuming that the user continuously slides two screens without clicking any product, it proves that the exposed products can not meet the needs of the user, and at this time, the label words with commonality can be extracted from the exposed products as negative label words (for example, most of the exposed products have a certain label word, and the label word belongs to a negative label word, and so on). After that, the user can be avoided from being recommended with the negative label words.

[0117] In the embodiment of the present application, the specific label words can include label words corresponding to the attributes of the goods determined by the goods information service system itself, and can also include label words mined from the user published content. That is, the information in the goods information service system can only determine the label words related to the attributes of the user, including the label words in the dimensions of style, material, and style (usually provided by the merchant or according to the information in the system), and the label words about the specific goods suitable for use in what occasion, whether it has a more fashionable style description, whether it is the current hot topic or trend, whether it is recommended by a celebrity, whether it is of interest to many people, etc. are difficult to obtain directly from the goods information service system. Therefore, in the embodiment of the present application, the user published content related to the goods can be collected from at least one content publishing system, and the specific information collected can include the publisher information of the specific content, and the specific content subject can be a piece of text and image, or a piece of short video, live broadcast, etc. Then, the label words are mined from the analysis of the user published content. Since the user published content is an introduction or recommendation of the goods in the use state, the specific label words mined can be related to the scene, style, trend, celebrity recommendation, and many people "grass-roots", etc. Although these label words belong to more subjective description information, such label words can reflect the characteristics of the goods from the aspects of practicality and authority, and therefore, are more consistent with the expression of the practicality of the searcher user.

[0118] It should be noted that the goods can be provided with label words of style and function according to the information in the goods information service system, but the label words of style and function added to the specific goods by the information in the goods information service system can not be comprehensive or not fashionable enough. For example, some goods belong to a certain style but are not added with the corresponding label; or some fashionable styles can not be obtained by the goods information service system in time, but can be known or discussed by some users who are at the forefront of fashion in the content publishing system, etc. Therefore, the style label words can be mined from the user published content to supplement the style label words in the goods information service system, or to add style label words to more goods, etc.

[0119] Therefore, in the embodiments of the present application, the collection of user published content related to goods can be performed in advance, and then some label words can be mined through algorithm analysis and the like. Since the label words correspond to specific goods, after the label words are determined, the information of the goods corresponding to the label words in the user published content can be used, for example, the specific user published content can include pictures of the goods, or videos or text information for explaining the goods, including the description of the brand, style and the like of the goods, so that the pictures, videos and text information and the like can be used as the information of the goods. Then, the goods matching the above information of the goods can be determined from the goods information database of the goods information service system, and the corresponding label words can be used to add labels to the matching goods. For example, a label word is mined from some user published content, and the goods involved in the user published content can be determined as the goods related to the label word, and then the goods matching the goods in the goods information database (which can be one or more) can be determined, and the label word can be used to add labels to the matching goods. For example, the pictures of the goods involved in the user published content related to a label word can be used to compare the similarity with the pictures of specific goods in the goods information database to find the same or similar goods. Or, the brand, style and the like of the goods can be directly included in some user published content, and more accurate goods matching can be performed in the goods information database based on the information, and the like.

[0120] As described above, the label word mining can be performed according to different goods categories, that is, the label word mining can be performed for some goods categories (which can be some large categories) in the specific goods information service system. In this way, the respective label word sets can be obtained for different goods categories. In addition, in the same goods category, a plurality of label words in a plurality of label dimensions can be included, for example, the label words in the scene and style dimensions can be included in the “sweater” category. And in the “jacket” category, the scene and style can be fixed, and the label word mining can be performed in the function (warmth, waterproofness and the like) dimension, and the like.

[0121] In the above manner, the label dimensions in the specific goods category and the label words included in each label dimension can be determined. The label dimensions required in the specific goods category can be determined through operation configuration and the like, and the label words in the specific label dimension can be determined by mining from the collected user published content. Or, the label words can be mined first, and then the label words can be clustered upward to obtain the specific label dimension, and the like.

[0122] Among them, as described above, the specific excavated label word can include: a classification label word used to subdivide the goods under the commodity category, at this time, when determining the target label word according to the collected positive feedback and / or negative feedback operation behavior information, the classification label word can be included to assist in improving the accuracy of the search word.

[0123] Specifically, the above-mentioned classification label word can also have multiple, for example, in one way, it can include: obtaining the description word from the user production content related to the goods, which describes the characteristics of the goods in use, and generating a label word after clustering. That is, the specific label word is more described by the language habits used by users in daily life, for example, "date fashion" and the like, rather than the traditional way of seemingly objective but lacking practicality.

[0124] Among them, the specific commodity category related classification label word can correspond to multiple label dimensions, so as to respectively subdivide the goods under the commodity category from multiple label dimensions. For example, specifically, the label dimension includes: scene dimension, the classification label word under this dimension includes: a label word used to describe the use scene suitable for the goods. For example, regarding the "sweater" category, under the scene dimension, it can be subdivided into "date fashion", "high-end fashion", etc. In addition, the specific label dimension can also include style dimension, the classification label word under this dimension includes: a label word used to describe the style to which the goods belong. For example, regarding the "sweater" category, under the style dimension, it can be subdivided into "comfortable and lazy ins style", "sports function style", etc. Of course, it can also include label words of other dimensions such as performance, material, element, etc. Although some label words can be obtained through the information inside the commodity information service system under these dimensions, but through the embodiment of the present application, more and richer label words under the specific dimension can be excavated from the user production content.

[0125] Specifically, after adding the corresponding label word to the matched goods, the preference information of each people group to the classification label word can also be determined through the historical behavior data of the consumer users corresponding to the goods in the commodity information service system, and the people group label information to which the consumer users belong. For example, the consumer users in a certain people group can be counted to purchase or browse the most goods, and then the label words preferred by the people group can be counted according to the label words carried by these goods, etc. In this way, when providing the classification label word corresponding to the target commodity category to the current searcher user, the target people group label to which the current searcher user belongs can be determined first, and then the target label word can be determined according to the label word possessed by the goods in the search result, the label word preferred by the target people group, and the operation behavior information.

[0126] In addition to the aforementioned classification label words, topic discussion content related to the product category can be determined from relevant user-generated content, and label words belonging to a discussion hotspot or trend can be mined from the topic discussion content. Specifically, content publishing systems often have topics such as "Daily Outfit", "Today's Style", and the like. Each topic can attract some users to generate content, and different users can also discuss specific topics, and the like. Therefore, label words belonging to a discussion hotspot or trend can be mined, for example, under the "coat" category, the latest trend words that can be mined include "tassels", "checkerboard", and the like.

[0127] In providing the recommended classification label words, the target label words that need to be recommended can be determined according to the operation behavior information of the positive feedback and / or negative feedback generated by the searcher user, and the timing of displaying the classification label words can be determined during the process in which the searcher browses the search results. For example, in the process of providing search results according to the product search request, if the searcher user continuously slides multiple screens (for example, two screens) of the search result page without clicking, it proves that the current search results may not meet the user's needs, at which time the determination and display of the target label words can be triggered.

[0128] Alternatively, if the searcher user selects a product link from the search results, clicks to view the product detail page, and then returns to the search result page, the determination and display of the target label words can also be triggered at this time, and the like.

[0129] That is, in the embodiments of the present application, instead of providing multiple label words of a navigation nature at the top of the search result page, the label word that needs to be recommended can be dynamically determined according to the operation of the user on the search result page, so as to present the effect of "accompanying search".

[0130] In specific implementation, since the purpose of providing the target label word in the embodiments of the present application is to help the user obtain more accurate search results, rather than for navigation purposes, a single target label word can be provided each time for the searcher user to select. For example, as shown in 31 in Figure 3 , the label word "date fashion" under the scene dimension can be provided, so that the user only needs to determine whether to select the label word to supplement to the keyword, rather than navigating the user by displaying multiple label words.

[0131] In the case of displaying only a single target label word each time, the user can also select the label word in a more convenient manner. For example, the user can directly click the keyword in the state of displaying a single label word, as a confirmation of the selection of the label word. Correspondingly, for the client, if the operation of the user clicking the label word is received, the classification label word can be added to the keyword input control (or search box) as a supplement to the current keyword, and the search result is refreshed. When the search result is refreshed, the goods can be filtered and / or reordered.

[0132] In the display of the recommended target label word, it can also be displayed within a certain threshold time period, and in order to avoid being ignored by the user, an animation effect can also be presented when displaying such target label. For example, the target label word can enter from one side of the screen, float to the middle of the screen and stay for a few seconds; then if the user clicks the label word, the label can be added to the search box, and the search result page is refreshed; if the user does not click, it will float away, similar to the effect of a bullet screen. In addition, if the user does not select a target label word, after a period of time, the user can provide a new recommended label word according to the new operation behavior of the user and / or the operation behavior generated in the current search process. The newly recommended label word can belong to the same dimension as the previously recommended label word, or it can be a different dimension. Of course, when recommending a new target label word, the label word that has been recommended but not executed by the user can be determined as a negative label word, and then the negative label word is removed from the multiple label words possessed by the search result of the goods, and the recommended label word is selected from the remaining label words. That is, in each recommendation, the target label word can be differentiated in the display, for example, if the user still does not click after recommending "high collar", the next recommended label word can be "round collar", or it can be a label word in other dimensions, and so on. In order to avoid too much interference to the normal browsing of the user, the interval time between two label word displays can be no less than a certain threshold, for example, 5S, etc.

[0133] By providing the user with a target label word for further subdividing the goods category corresponding to the current keyword during the user's browsing of the search result, and the target label word can include a label word mined from the user's published content, the user can be provided with more practical auxiliary information, thereby effectively helping the user to improve the accuracy of the search result.

[0134] In addition to providing the target label words with the classification property as described above, exclusive label words provided for specific goods can also be provided. That is, the specific label words can also include exclusive label words corresponding to individual goods obtained from the user-generated content related to the goods, including label words related to celebrity same goods or the number of people interested. For example, as described above, assuming that a celebrity recommends a certain good from a user's published content, the system can determine the goods related to the good in the goods information service system and add exclusive labels such as "recommended by a certain star" to the goods. Or, if a certain good is found to be "grassed" by a large number of people in the content publishing system, the system can determine the goods related to the good in the goods information service system and add exclusive label words such as "a certain number of people have been grassed" to the goods. For such exclusive label words, the system can also display the exclusive label words related to celebrity same goods or the number of people interested corresponding to the goods when displaying the goods in the search results to assist the searcher user in making a decision on the goods from the search results.

[0135] To mine label words related to celebrity same goods, etc., the system can also obtain the list information of "stars" or "big Vs" in the content publishing system in advance by the operator, etc., and configure it in the algorithm, so that the specific algorithm can determine which accounts in the content publishing system belong to the accounts corresponding to the celebrity, and then determine whether a celebrity has recommended a certain good, etc. through the user-generated content associated with such accounts.

[0136] That is, the search assistance information provided to the user can be divided into two ways. One is to display the target label words for subdividing the goods category in the search result page in the form of bullet screen, etc., and if the user selects, it can be added to the current keyword to help the user narrow down the search range and improve the accuracy of the search results. The other way is to directly display the exclusive label words associated with specific goods in the search results, including label words related to celebrity recommendations or many people interested, to reflect the authority and help the user make a decision on the goods in the search results.

[0137] It should be noted that in the process of providing the specific label words, the system can also provide the identification information of the content publishing system corresponding to the specific label words. For example, for the exclusive label words related to the number of people interested in a certain good, the system can display "a certain number of people have been grassed in a certain content publishing system" based on "a certain number of people have been grassed", etc.

[0138] In addition, real-time feature extraction can be performed on the non-clicked commodity of the user, and the extracted features can be analyzed. When a search request initiated by the user is subsequently recalled, the commodities with the same or similar features can be given a lower weight.

[0139] In summary, through the embodiments of the present application, in the process of providing commodity search results based on the current search keywords, the user's operation behavior information related to positive feedback and / or negative feedback can be collected, such as the user's positive feedback behavior of clicking to view the details of a certain commodity, or the negative feedback behavior of continuous sliding without clicking, etc. Then, according to the operation behavior information, the target label word can be selected from the label words possessed by the commodities in the commodity search results, and recommended to the user, so as to guide the user to provide more accurate search keywords through the target label word. That is, in the embodiments of the present application, some optional label words can also be recommended to the user to guide the user to search more accurately, but the specific label words to be displayed are determined according to the positive feedback and / or negative feedback reflected by the user's operation behavior in the search process, so as to form a recommended label in the form of interaction, rather than simply listing multiple label words in the navigation bar. In other words, in the embodiments of the present application, the specific recommended label words are generated along with the user's operation behavior, and through the user's continuous positive feedback or negative feedback behavior, the label words that are more likely to meet the user's needs can be gradually recommended. Moreover, the user is allowed to choose whether to use the specific label word to achieve more accurate search, rather than automatically selecting in the background, so as to give the user the initiative to choose, so that the user can obtain a sense of participation, and the search results can also meet the user's needs.

[0140] Among them, the specific recommended label words can be pre-analyzed from the user's production content (which can include text and image content, short video, live content, etc.) and mined from the label words, so as to enrich the label words generated based on the attributes of the commodities. The label words mined in this way are closer to the user's feelings or experience summary of the commodities in actual life scenarios, and therefore can have the characteristics of being more practical, more comprehensive, and more reflecting the current popular trends, so as to help the user better express his search needs.

[0141] Embodiment Two

[0142] In the foregoing embodiment one, the target tag word is recommended to the user according to the operation behavior information related to the positive feedback and / or negative feedback generated by the user, mainly in the process of displaying the search result. The specific tag word recommended can include the tag word mined from the user published content. In this embodiment two, the mined tag word can also be directly used to recommend the tag word to the user, and after the user selects to use a specific tag word, the search can be re-initiated to update the search result. Specifically, refer to Figure 4 This embodiment two provides another method for processing the commodity search, which can specifically include:

[0143] S401: providing a commodity search result based on a current search keyword;

[0144] S402: providing a recommended tag word in the process of providing the search result, wherein the tag word is obtained by analyzing and mining the user published content related to the commodity information;

[0145] As described in the embodiment one, the tag word mining can be performed for different commodity categories according to the commodity category system in the commodity information system, and therefore, the specific tag word mined can have an association with a specific commodity category. The current keyword can also be related to a specific commodity category, and therefore, the specific tag word and the current search keyword can be associated through the commodity category. In this way, when the recommended tag word is provided, the target commodity category can be determined according to the current search keyword, and then the tag word to be recommended can be determined according to the tag word associated with the target commodity category.

[0146] In addition, the preference information of multiple people groups for the tag word associated with each commodity category can also be determined, that is, the preference of each people group for the tag word under a specific commodity category is determined, and therefore, the target people group to which the searcher user belongs can be determined first, and then the target tag word can be determined according to the tag word preferred by the target people group under the target commodity category.

[0147] S403: after the recommended tag word is selected, the current search keyword and the recommended tag word are combined as a new search keyword, and after the search is re-initiated and updated, a new commodity search result is provided.

[0148] In a specific implementation, the recommended tag words can be displayed within a certain threshold time period, for example, through a "floating" display in a bullet screen or other form. At this time, if the user does not select the recommended tag word within the threshold time period, a new tag word can be recommended to the user. For example, there may be multiple tag words related to the current search keyword, and other tag words can be replaced for recommendation.

[0149] For the parts not described in detail in the second embodiment, please refer to the description in the first embodiment, which will not be repeated here.

[0150] 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.).

[0151] Corresponding to the first embodiment, the present application embodiment also provides a processing device for product search, see Figure 5 , the apparatus may include:

[0152] The operation behavior information acquisition unit 501 is used to acquire operation behavior information related to positive feedback and / or negative feedback generated by users in the process of providing product search results based on the current search keyword;

[0153] a tag word determining unit 502 for determining a target tag word from the tag words of the products in the product search results according to the operation behavior information;

[0154] The tag word recommendation unit 503 is configured to recommend the target tag word to the user, so as to guide the user to provide more accurate search keywords through the target tag word.

[0155] The operation behavior information acquisition unit may be specifically used to:

[0156] Obtaining operational behavior information related to the positive feedback and / or negative feedback generated by the user on the product search results.

[0157] Specifically, the operation behavior information acquisition unit can be used to:

[0158] If the user performs an operation of viewing detailed information of a product in the product search results, the product is determined as a positive product that has received positive feedback from the user;

[0159] At this time, the label word determination unit can be specifically used to:

[0160] When returning to the commodity search result page, a target label word is selected from the label words possessed by the positive commodity as the recommended target label word.

[0161] Alternatively, the operation behavior information acquisition unit can be specifically configured to:

[0162] If the user continuously slides multiple screens on the search result page without a click operation, the commodities displayed by the sliding are determined as multiple negative commodities that obtain negative feedback from the user.

[0163] At this time, the label word determination unit can be specifically configured to:

[0164] The label word common to the multiple negative commodities is determined as a negative label word.

[0165] After the negative label word is removed from the multiple label words possessed by the commodities in the commodity search result, the recommended target label word is selected from the remaining label words.

[0166] In addition, the apparatus can further include:

[0167] The recommendation updating unit is configured to update the recommended target label word after obtaining new operation behavior information related to positive feedback and / or negative feedback.

[0168] Further, the apparatus can further include:

[0169] The new keyword generation unit is configured to combine the target label word and the current search keyword into a new search keyword after receiving a target operation performed by the user on the recommended target label word.

[0170] The search result refreshing unit is configured to refresh the commodity search result according to the new search keyword.

[0171] The recommended target label word is displayed within a threshold time period.

[0172] At this time, the apparatus can further include:

[0173] The re-recommendation unit is configured to provide the user with a recommended label word again according to new operation behavior of the user and / or operation behavior generated in the current search process if no target operation performed by the user on the recommended target label word is received within the threshold time period.

[0174] Further, the apparatus can further include:

[0175] The negative label word determination unit is configured to determine a label word that has been recommended but on which no target operation is performed by the user as a negative label word.

[0176] The label word reselection unit is configured to select the recommended label word from the remaining label words after removing the negative label word from the plurality of label words possessed by the commodity search result when re-providing the recommended label word for the user.

[0177] The recommended target label word is displayed in the form of a floating subtitle within a threshold time period.

[0178] The recommended single label word is displayed within the same threshold time period.

[0179] The target label word includes a label word obtained by analyzing and mining user-generated content related to the product description, and adding a label to related commodities in the commodity information service system according to product information associated with the user-generated content, wherein the user-generated content includes text, short video and / or live content.

[0180] The target label word includes a label word obtained by analyzing and mining user-generated content related to the product description, and adding a label to related commodities in the commodity information service system according to product information associated with the user-generated content, wherein the user-generated content includes text, short video and / or live content.

[0181] And / or,

[0182] The related topic discussion content is determined from the user-generated content related to the product, and a label word belonging to a discussion hot spot or trend is mined from the topic discussion content.

[0183] In addition, the label word obtained by analyzing and mining the user-generated content can also include a dedicated label word corresponding to a single commodity, wherein the dedicated label word includes a label word related to a celebrity or a number of people interested in the same product.

[0184] At this time, the device can further include:

[0185] The dedicated label word providing unit is configured to display the dedicated label word related to a celebrity or a number of people interested in the same product corresponding to the commodity search result when displaying the commodity search result, to assist the user in making a selection decision from the commodity search result.

[0186] In addition, the device can further include:

[0187] The preference information determination unit is configured to determine the preference information of each group for the label word according to the user historical behavior data corresponding to the commodity, the label word possessed by the commodity, and the group label information to which the user belongs.

[0188] The label word determination unit can be specifically configured to:

[0189] Determine a target group to which the current search user belongs.

[0190] According to the label words possessed by the commodities in the commodity search result, the label words preferred by the target group, and the operation behavior information, the target label word is determined.

[0191] According to the embodiment two, the application embodiment further provides a processing device for commodity search, referring to Figure 6 The device can include:

[0192] A search result providing unit 601 is configured to provide a commodity search result based on a current search keyword;

[0193] A label word recommending unit 602 is configured to provide a recommended label word in the process of providing the search result, wherein the label word is obtained by analyzing and mining user production content related to commodity information;

[0194] A re-search unit 603 is configured to combine the current search keyword and the recommended label word as a new search keyword after the recommended label word is selected, and provide a new commodity search result after re-initiating the search.

[0195] The label word is associated with a commodity category, and a plurality of label words are included under the same commodity category.

[0196] The label word recommending unit can be specifically configured to:

[0197] Determine a target commodity category according to the current search keyword;

[0198] Determine a label word to be recommended according to the label words associated with the target commodity category.

[0199] In specific implementation, the device can further include:

[0200] A preference information determining unit is configured to determine preference information of each group of people for label words associated with each commodity category.

[0201] At this time, the label word recommending unit can be specifically configured to:

[0202] Determine a target group of people to which a searcher belongs;

[0203] Determine the target label word according to the label words preferred by the target group of people under the target commodity category.

[0204] Specifically, the recommended label word is displayed within a threshold time period.

[0205] At this time, the device can further include:

[0206] The re-recommendation unit is configured to recommend a new label word to the user if a selection operation on the recommended label word is not received from the user within the threshold time period.

[0207] In addition, the embodiments of the present application further provide a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the steps of the method according to any one of the preceding method embodiments.

[0208] and an electronic device comprising:

[0209] one or more processors; and

[0210] a memory associated with the one or more processors, the memory configured to store program instructions that, when executed by the one or more processors, perform the steps of the method according to any one of the preceding method embodiments.

[0211] wherein, Figure 7 An exemplary shows the architecture of an electronic device, for example, the device 700 can be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, an aircraft, etc.

[0212] Referring to Figure 7 , the device 700 can include one or more of the following components: a processing component 702, a memory 704, a power supply component 706, a multimedia component 708, an audio component 710, an input / output (I / O) interface 712, a sensor component 714, and a communication component 716.

[0213] The processing component 702 usually controls the overall operation of the device 700, such as operations associated with displaying, making phone calls, data communications, camera operations and recording operations. The processing component 702 can include one or more processors 720 to execute instructions to complete all or part of the steps of the methods provided by the technical solutions of the present disclosure. In addition, the processing component 702 can include one or more modules to facilitate interaction between the processing component 702 and other components. For example, the processing component 702 can include a multimedia module to facilitate the interaction between the multimedia component 708 and the processing component 702.

[0214] The memory 704 is configured to store various types of data to support operations of the device 700. Examples of such data include instructions for any application or methods operating on the device 700, contact data, phonebook data, messages, pictures, videos, and the like. The memory 704 can be implemented by any type of volatile or nonvolatile storage devices or a combination thereof such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read only memory (PROM), read only memory (ROM), magnetic memory, flash memory, magnetic disk, or optical disk.

[0215] The power component 706 provides power to the various components of the device 700. The power component 706 can include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 700.

[0216] The multimedia component 708 includes a screen providing an output interface between the device 700 and a user. In some embodiments, the screen can include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense a touch, a slide, and a gesture on the touch panel. The touch sensor can not only sense a boundary of a touching or a sliding action, but also detect duration and pressure related to the touching or sliding action. In some embodiments, the multimedia component 708 includes a front camera and / or a back camera. The front camera and / or the back camera can receive external multimedia data when the device 700 is in an operating mode, such as a shooting mode or a video mode. Each of the front and back cameras can be a fixed optical lens system or have a focal length and optical zoom capability.

[0217] The audio component 710 is configured to output and / or input an audio signal. For example, the audio component 710 includes a microphone (MIC) configured to receive an external audio signal when the device 700 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signal can be further stored in the memory 704 or transmitted via the communication component 716. In some embodiments, the audio component 710 also includes a speaker for outputting an audio signal.

[0218] The I / O interface 712 provides an interface between the processing component 702 and peripheral interface modules, which can be a keyboard, a click wheel, a button, and the like. The buttons can include, but are not limited to, a home button, a volume button, a start button, and a lock button.

[0219] The sensor component 714 includes one or more sensors for providing status assessments for various aspects of the device 700. For example, the sensor component 714 can detect an open / closed position of the device 700, relative positioning of components, such as a display and keypad of the device 700, changes in position of the device 700 or a component of the device 700, presence or absence of user contact with the device 700, changes in orientation of the device 700 or acceleration / deceleration, and temperature changes of the device 700. The sensor component 714 can include proximity sensor(s) configured to detect presence of nearby objects without any physical contact. The sensor component 714 can also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor component 714 can also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0220] The communication component 716 is configured to facilitate wired or wireless communication between the device 700 and another device. The device 700 can access a wireless network based on a communication standard, such as WiFi, or a 2G, 3G, 4G / LTE, 5G, or the like cellular communication network. In an example embodiment, the communication component 716 receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an example embodiment, the communication component 716 further includes a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on Radio Frequency Identification (RFID) techniques, infrared data association (IrDA) techniques, ultra-wideband (UWB) techniques, Bluetooth (BT) techniques, and other techniques.

[0221] In an example embodiment, the device 700 can be implemented using one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, micro-controllers, microprocessors, or other electronic components, for performing the above-described methods.

[0222] In an example embodiment, a non-transitory computer-readable storage medium, such as the memory 704 including instructions, is also provided, which can be executed by the processor 720 of the device 700 to complete the methods provided by the techniques of this disclosure. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disc, and an optical data storage device, and the like.

[0223] Those skilled in the art can clearly understand the application by the description of the above embodiments that the application can be implemented by means of software and the necessary universal hardware platforms. Based on such an understanding, the technical solutions of the application can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, and the like) execute the methods described in each embodiment or some parts of the embodiments of the application.

[0224] The various embodiments in the specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the system or system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments. The above-described system and system embodiments are merely illustrative, and the units described as separate components can be or can not be physically separated, and the components displayed as units can be or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. According to the actual needs, some or all of the modules can be selected to achieve the purpose of the embodiment. Those skilled in the art can understand and implement without creative labor.

[0225] The above provides a detailed description of the product search processing method and the electronic device provided by the application. The principle and implementation manner of the application are described by applying specific examples. The above embodiment is only used to help understand the method and the core idea of the application. Meanwhile, for those skilled in the art, according to the idea of the application, the specific implementation manner and application range can be changed. In summary, the content of the specification should not be understood as a limitation of the application.

Claims

1. A method for processing product search, characterized in that: include: In the process of providing product search results based on the current search keyword, obtaining operational behavior information related to positive feedback and / or negative feedback generated by users on the product search results; Determining target tag words from the tag words of the products in the product search results based on the operation behavior information; wherein the tag words include: obtaining descriptive words describing the characteristics of the product in use from user-generated content related to product descriptions in advance in at least one content publishing system, clustering the descriptive words to generate tag words for expressing the user's practical demands, and adding corresponding target tag words to products matching the target products in the product information service system based on the target product information associated with the user-generated content; the user-generated content includes pictures, text, short videos, and / or live broadcast content; The target tag word is recommended to the user so as to guide the user to add the target tag word to the search keyword, so as to re-initiate the search based on the new search keyword added by the user to obtain refreshed product search results.

2. The method according to claim 1, characterized in that The obtaining of the user's operational behavior information related to the positive feedback and / or negative feedback generated by the user on the product search results includes: If the user performs an operation of viewing detailed information of a product in the product search results, the product is determined as a positive product that has received positive feedback from the user; The step of determining a target tag word from the tag words of the products in the product search results according to the operation behavior information includes: When returning to the product search result page, a target label word is determined from the label words of the positive product and used as the recommended target label word.

3. The method according to claim 1, characterized in that The obtaining of the user's operational behavior information related to the positive feedback and / or negative feedback generated by the user on the product search results includes: If the user swipes multiple times on the search results page without clicking, the products displayed by the swipe are determined to be multiple negative products that have received negative feedback from the user; The step of determining a target tag word from the tag words of the products in the product search results according to the operation behavior information includes: Determining the common label words of the multiple negative products as negative label words; After removing the negative label words from a plurality of label words of the product in the product search results, the recommended target label word is determined from the remaining label words.

4. The method according to claim 1, wherein The recommended target tag word is displayed within a threshold time period; The method further comprises: If the user does not perform a target operation on the recommended target tag word within the threshold time period, the recommended tag word is provided to the user again based on the user's new operation behavior and / or the operation behavior that has occurred during the current search process.

5. The method according to claim 4, characterized in that Also includes: Determine the recommended label words for which the user has not performed the target operation as negative label words; When re-providing the user with recommended label words, after removing the negative label words from the multiple label words in the product search results, the recommended label words are selected from the remaining label words.

6. The method according to any one of claims 1 to 5, characterized in that Also includes: Based on the user's historical behavior data corresponding to the product, the product's label words, and the user's group label information, determine the preference information of each group for the label words; The step of determining a target tag word from the tag words of the products in the product search results according to the operation behavior information includes: Determine the target group to which the current search user belongs; The target label word is determined according to the label words of the products in the product search results, the label words preferred by the target group, and the operation behavior information.

7. A method for processing product search, characterized in that: include: Provide product search results based on the current search keywords; In the process of providing the search results, recommended tag words are provided, wherein the tag words include: pre-acquired from at least one content publishing system user-generated content related to product descriptions, descriptive words describing characteristics of the product under usage, clustering the descriptive words to generate tag words used to express the user's practical demands, and adding corresponding tag words to products in the product information service system that match the target product based on target product information associated with the user-generated content; the user-generated content includes pictures, text, short videos, and / or live broadcast content; After the recommended tag word is selected, the recommended The tag word is added to the current search keyword to re-initiate the search based on the new search keyword added by the user to obtain refreshed product search results.

8. The method according to claim 7, characterized in that The tag words are associated with the product category, and the same product category includes multiple tag words; The recommended tag words include: Determine the target product category based on the current search keyword; Determine the tag words to be recommended based on the tag words associated with the target product category.

9. The method according to claim 8, characterized in that Also includes: Determine the preferences of different groups of people for the labels associated with each product category; The step of determining the tag words to be recommended based on the tag words associated with the target product category includes: Determine the target demographic to which the searcher belongs; The target label word is determined according to the label words preferred by the target group under the target product category.

10. 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 according to any one of claims 1 to 9 are implemented.

11. 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 according to any one of claims 1 to 9.

Images