Idle items management method and system

Abstract

The present application relates to the technical field of idle item information management, and in particular to an idle item management method and system, wherein the method includes a database construction step, a label construction step, an information acquisition step, an information processing step, an information interaction step, a push step, a feedback step, and a label supplement step; wherein the system includes a database, a label construction module, an information acquisition module, an information processing module, an information interaction module, a push module, a feedback module, and a label supplement module. The present application has the effect of improving the accuracy of idle item classification management and improving the efficiency of resource reuse.

Description

Technical Field

[0001] The present application relates to the technical field of idle item information management, and in particular to an idle item management method and system. Background Art

[0002] At present, with the development of social economy and the improvement of people's living standards, consumers are upgrading products at a faster speed. After many people buy new products, old products may be regarded as idle items. These items may not be needed by their owners temporarily, or they are no longer used frequently for some reasons, but they are functionally intact and still have use value. In order to realize the value of idle items, idle item trading activities are generated. Through trading, these items can be recirculated, avoiding the waste of resources and realizing the maximum utilization of resources.

[0003] With the development of information data, transactions of idle items are generally conducted through trading platforms. However, a large amount of idle item data is stored in the trading platform. The same type of products may have different specifications, and the classification labels are not detailed, which can easily lead to confusion in management and classification. For consumers, they need to select items that meet their requirements from a large amount of redundant information on similar items, which wastes a lot of time and leads to inefficient resource reuse. Summary of the invention

[0004] In order to improve the accuracy of idle item classification management and increase resource reuse efficiency, the present application provides an idle item management method and system.

[0005] In the first aspect, the present application provides a method for managing idle items, which adopts the following technical solution:

[0006] A method for managing idle items comprises the following steps:

[0007] Build a database: Establish a database that stores item category information tables, idle item information, and idle item labels;

[0008] Label construction: Generate idle item labels based on the item category information table, where idle item labels include category labels, basic labels and supplementary labels;

[0009] Information acquisition: obtain the idle items information registered by the publishing user;

[0010] Information processing: including storage segmentation step and the first labeling step;

[0011] Storage segmentation: extract the registered keywords according to the idle items information registered by the publishing user, and form the first keyword information;

[0012] First labeling: matching idle item labels according to the first keyword information, associating the matched idle item labels with the idle items, and storing the associated information in a database;

[0013] Information interaction: including demand acquisition step, demand segmentation step and demand matching step;

[0014] Demand acquisition: obtain query demand information;

[0015] Demand segmentation: extract query keywords based on query demand information to form second keyword information;

[0016] Demand matching: matching idle item tags according to the second keyword information, and determining whether there are idle items that meet the matching tags according to the matched idle item tags combined with the associated information. If so, executing the push step, otherwise, executing the feedback step;

[0017] Push: Push the idle item information to the querying user;

[0018] Feedback: Prompt the querying user that there is no idle item in demand;

[0019] Label supplement: including data statistics steps and optimization feedback steps;

[0020] Data statistics: Query the pushed items that are filtered out by users, obtain the reasons for the removal, and generate supplementary labels;

[0021] Optimization feedback: Update the generated supplementary tags to the idle item tags, establish the association information between the supplementary tags and the idle items, and add additional tags to the idle items for subsequent inquiries by other query users.

[0022] By adopting the above technical solution, by building a database and constructing a detailed idle item label system (including category labels, basic labels and supplementary labels), the method can systematically organize and store idle item information, making the management of item information more orderly and efficient; in the information processing stage, by storing word segmentation and the first labeling step, the idle item information registered by the user is converted into standardized keyword information, and matched with the predefined idle item labels, thereby realizing fast and accurate classification and indexing of idle items, which greatly improves the retrieval efficiency and accuracy of subsequent queries; in the information interaction stage, the method can accurately capture the needs of the querying user through the steps of demand acquisition, demand segmentation, demand matching, etc., and quickly push the idle item information that meets the needs. At the same time, through the label supplement mechanism, the idle item label system is continuously optimized and improved, so that the push results are closer to the actual needs of the user. This improves the user experience and can optimize the push data of post-query users based on the feedback of pre-query users, making it more in line with the actual needs of query users, improving the accuracy of idle item classification management, improving resource reuse efficiency, and improving disposal efficiency, thereby promoting the completion of transactions; through efficient information management and precise push mechanism, this method helps to promote the effective circulation of idle items, users can more easily find the items they need, and the owners of idle items can also find potential buyers or demanders more quickly, thereby achieving optimal allocation and reuse of resources; the method also includes label supplement and optimization feedback steps, and continuously generates and updates supplementary labels by statistically querying users' feedback and exclusion reasons to achieve continuous optimization of the idle item label system; this dynamic optimization mechanism enables the entire system to continuously adapt to new needs and changes and maintain its efficiency and accuracy.

[0023] Optionally, a matching calculation step and a grading and screening step are further provided between the demand matching step and the feedback step;

[0024] Matching calculation: Calculate the similarity between the second keyword information and the idle item label, and calculate their sum value to obtain the idle item matching degree;

[0025] Grading screening: Determine whether the idle item matching degree is greater than the set push matching value. If so, execute the summary step; otherwise, execute the feedback step;

[0026] Summarize: Summarize the idle items whose matching degree is greater than the set push matching value, and then perform the sorting step;

[0027] Sorting: Sort the idle item information in descending order of matching degree, and then execute the push step.

[0028] By adopting the above technical solution, by calculating the similarity between the second keyword information and the idle item label and summing them up to obtain the idle item matching degree, this step can more finely measure the correlation between the query demand and the idle items. Compared with simple label matching, this method can better capture the subtle differences in the query demand, thereby improving the matching accuracy; the introduction of the matching calculation step enables the system to automatically calculate the matching degree according to the query demand and perform subsequent processing according to the matching degree. This automated processing method enhances the intelligence of the system and enables it to better adapt to the query needs of different users; by setting the push matching value, the system can filter out idle items with a matching degree higher than this value for push; this The screening method can ensure that the items pushed to users are highly relevant to their needs, thereby improving the accuracy and effectiveness of the push; the grading and screening step can effectively filter out idle items with low matching degrees and avoid pushing these items to users; this can not only reduce the screening burden of users, but also improve the user's satisfaction with the push results; by reducing invalid pushes, the system can more efficiently use computing resources and network resources and improve the overall operating efficiency; at the same time, since the push results are more accurate, users are also more likely to find satisfactory items, thereby improving the success rate of transactions; the idle items with a matching degree greater than the set push matching value are aggregated, which helps the system to manage these items in a unified manner. This not only simplifies the push process, but also facilitates subsequent data analysis and optimization; the aggregated push results are more centralized and orderly, and users can more conveniently browse and select items of interest. This way of improving user experience helps to increase user stickiness and activity. This method, through the newly added matching calculation step, grading and screening step, together constitutes a more complete and efficient demand matching and push mechanism. This mechanism not only improves the matching accuracy and push accuracy, but also enhances the intelligence and efficiency of the system, providing users with a better and more convenient idle item trading experience; the sorting step enables the system to push the most relevant items to users more efficiently. Compared with random or disordered push methods, this orderly push strategy can attract users' attention more quickly and guide them to take further actions; the sorted push results help to expose idle items with high matching degrees to potential buyers more quickly, thereby promoting the circulation and reuse of items. This not only helps to optimize the allocation of resources, but also may bring about economic benefits.

[0029] Optionally, the sorting step further includes obtaining the registration time of the idle items, and sorting the idle items with the same matching degree according to the registration time, wherein the sorting is performed in ascending order of the registration time.

[0030] By adopting the above technical solution, newly released idle items will be displayed to the querying users first by sorting them according to the registration time. This helps to promote the rapid circulation of new idle items, because newly released items tend to attract more attention, and sellers may be more willing to accept reasonable prices; idle items that have not been sold for a long time (i.e. items with a long registration time) will be ranked later, which helps to reduce the retention time of these items in the system. By automatically adjusting the sorting, users are encouraged to update or reprice these items to attract more potential buyers; when users browse, they first see newly released items, which can bring a sense of freshness and increase the browsing interest and participation of querying users; for querying users who frequently visit the system, sorting by registration time helps reduce the possibility of them repeatedly viewing items they have already seen, thereby improving browsing efficiency and browsing coverage, while facilitating item comparison and obtaining the desired idle items faster.

[0031] Optionally, after the tag building step, a language library building step and a semantic expansion step are also provided;

[0032] Language library construction: Generate multi-language information based on the item category information table and establish a language library;

[0033] Semantic expansion: Based on the idle item tags and language library, the idle item tags are expanded, and subsidiary tags of the category tags are generated and updated to the idle item tags.

[0034] By adopting the above technical solution, by generating multi-language information according to the item category information table and establishing a language library, the system can support tag management and query in a multi-language environment. This greatly enhances the internationalization capability of the system, allowing users from different countries and regions to browse and search for idle items in their familiar languages, thereby expanding the user base and market coverage; based on the idle item tags and language library, the system can conduct in-depth semantic analysis of the tags and expand the generation of subsidiary tags; this semantic expansion not only enriches the tag system, but also enhances the system's ability to understand the user's query intention; through the subsidiary tags, the system can more accurately capture the potential needs in the user's query, thereby providing more accurate matching results; after semantic expansion, the tag set of idle items has been expanded, which enables the system to have more reference information when matching needs. Therefore, the matching accuracy will be significantly improved, and the system can more accurately find idle items that match the user's query needs. At the same time, due to the increase in the number of tags, the system may adopt more efficient algorithms or strategies when performing matching calculations, thereby improving matching efficiency; through language library construction and semantic expansion, the system can provide users with a more friendly and personalized query experience. Users can use the language they are familiar with to query and obtain more accurate and rich matching results. This optimization not only improves user satisfaction and loyalty, but also helps to enhance the brand image and market competitiveness of the entire platform; the subsidiary tags generated after semantic expansion provide the system with more data dimensions, which can be used for subsequent data analysis and optimization. Through statistics and analysis of tag usage, the system can find out which tags are more popular with users and which tags have poor matching effects, and optimize the tag system accordingly. At the same time, these data can also be used to improve recommendation algorithms, improve the quality of search result ranking, and so on; and for users with relatively poor Mandarin and unclear query needs, subsidiary tags can be generated through language expansion, so that the actual needs of the above users can be clearly represented, thereby improving matching accuracy and universality, which can promote the completion of transactions.

[0035] Optionally, a directional matching step is provided before the demand matching step;

[0036] Directed matching: Based on the query user information, determine the language preference, circle the range of idle item tags in the database, and then perform the demand matching step.

[0037] By adopting the above technical solution, the directional matching step first defines a relatively small range of idle item labels in a huge database based on the user's language preference and possible points of interest. This can significantly reduce the amount of data that needs to be processed in the subsequent demand matching step, thereby improving the overall matching efficiency; due to the reduction in the search range, the computing resources required by the system for matching calculations will also be reduced accordingly. This helps to improve the response speed of the system and reduce operating costs; through the directional matching step, the system can more accurately capture the user's query intentions and points of interest. This helps to recommend idle items that better meet the user's needs in the demand matching step, thereby improving the accuracy of matching; without the directional matching step, the system may include a large number of idle items that are irrelevant to the user's query intentions in the matching range. The directional matching step can effectively avoid this situation and ensure that the user sees results that are closely related to their query needs; the directional matching step provides the system with a basis for personalized recommendations. By filtering according to the user's language preference and interest points, the system can present idle item recommendations that are more in line with their personal preferences, thereby improving user satisfaction; since the targeted matching step has filtered out some irrelevant results for the user, the time and energy required for the user to browse and select will be reduced accordingly. This helps to improve the user's overall experience; the user information collected in the targeted matching step can be used to enrich the user portrait. By analyzing information such as user language preference and interest points, the system can have a deeper understanding of user needs and preferences, providing strong support for subsequent optimization and improvement; based on the results of the targeted matching step and the user's actual behavior data (such as clicks, browsing, purchases, etc.), the system can continuously optimize its recommendation algorithm. Through continuous iteration and optimization, the system can provide users with more accurate and personalized recommendation services; through the setting of targeted matching, it can match the top line in line with the user's real expression needs, reduce the situation where the same word has multiple semantics, improve the accuracy of idle item recommendations, and improve user experience.

[0038] Optionally, after the optimization feedback step, a conflict determination step is also included;

[0039] Conflict determination: determine whether there is a conflict between the supplementary tag and the existing tag in the association information between the idle items and the idle item tags. If so, execute the tag counting step;

[0040] Tag statistics: Count the number of query users defined as the supplementary tag, and determine whether the number of query users defined as the supplementary tag is greater than the set quantity limit. If so, perform the replacement step;

[0041] Replace: Delete the relationship between the conflicting existing tags and the corresponding idle items.

[0042] By adopting the above technical solution, the system can timely discover and solve the inconsistency problem in the label system by judging whether the supplementary label conflicts with the existing label. This conflict may be caused by the ambiguity, overlap or misunderstanding of the label definition. If it is not handled, it may mislead user queries and item matching. Through conflict discrimination, the system can ensure that each label represents a clear and independent concept, thereby improving the accuracy and consistency of the label system; by counting the number of query users defined as the same supplementary label, the system can evaluate the popularity and representativeness of the label. When the number is greater than the set quantity limit, it means that the label has been recognized and supported by more users, so it is necessary to further process it. This label optimization method based on user feedback can enhance user participation and satisfaction, because users feel that their opinions are valued and actually responded to; when a conflicting existing label is found, the system deletes the relationship between it and the corresponding idle item through the replacement step, and may replace it with a more suitable label. This replacement operation not only solves the label conflict problem, but also optimizes the structure and content of the label system. The optimized label system is clearer and more reasonable, which helps to improve the efficiency and accuracy of item matching. At the same time, the replacement step also provides the system with the ability to dynamically adjust and optimize the label system, so that it can better adapt to changes in market and user needs; through the implementation of conflict identification, label statistics and replacement steps, the system has accumulated a large amount of user feedback and label usage data. These data can be used for subsequent data analysis and decision support. For example, the system can analyze which labels are prone to conflict and which labels have high user recognition, thereby providing strong support for the optimization and improvement of the label system. At the same time, these data can also be used to evaluate the overall performance of the system and user satisfaction.

[0043] Optionally, after the data statistics step, an optimization and correction step is also provided;

[0044] Optimization and correction: Perform semantic inversion based on the supplementary tags, generate new second keyword information, and re-execute the demand matching step.

[0045] By adopting the above technical solution, the optimization and correction step generates new second keyword information by semantically negating the supplementary tag. This operation expands the semantic scope of the tag, allowing the tag to cover more query intent and context. For example, if the original tag is "fashion", semantic negation may generate relevant second keywords such as "non-fashion" or "retro", thereby helping the system to more accurately understand the deep-level needs of the user's query.

[0046] After generating the new second keyword information, the system re-executes the demand matching step. This step uses the updated tag system, including the original tags and the newly generated second keywords, to re-evaluate the match between the user query and the idle items. Because the tag system has been expanded and optimized, the system can more accurately identify idle items that match the user's query intent, thereby improving the accuracy and efficiency of matching.

[0047] By optimizing the correction steps, the system can better meet the diverse query needs of users. Whether it is a direct query or an indirectly related query intention, the system can provide more comprehensive and accurate matching results. This will significantly improve the satisfaction and convenience of user queries and enhance users' trust and reliance on the system.

[0048] The optimization and correction steps actually form a feedback loop, where the system generates additional tags based on user queries, evaluates the effectiveness of these tags through the statistical step, and then improves the tag system through the optimization and correction step. This cycle is constantly iterated, allowing the tag system to continue to adapt to market changes and changes in user needs, maintaining its timeliness and accuracy.

[0049] Optionally, a supplementary judgment step is provided between the data statistics step and the optimization feedback step;

[0050] Supplementary judgment: obtain the browsing times, rejection times and supplementary label generation times of the current idle items, and count the maximum number of times the same supplementary label is generated, and judge whether the maximum number of times the supplementary label is generated accounts for the number of rejection times of the querying user is greater than the set limit ratio. If so, execute the optimization feedback step, otherwise, execute the cycle determination step;

[0051] Cycle determination: Set the label update cycle to , statistical period The maximum number of times of generating supplementary labels and the number of times of query user rejection are determined, and whether the proportion of the maximum number of times of generating supplementary labels in the number of times of query user rejection is greater than or equal to the set limit ratio is determined. If so, the cycle maintenance step is executed, otherwise, the cycle correction step is executed;

[0052] Cycle maintenance: maintain the current tag update cycle and then execute the update step;

[0053] Cycle correction: The cycle is corrected according to the proportion of the maximum number of times that supplementary tags are generated in the number of times that query users are eliminated. The smaller the proportion, the greater the increase in the cycle. As the proportion increases, the trend increases, and then the supplementary judgment step is performed.

[0054] By adopting the above technical solution, the system can evaluate the effectiveness and user acceptance of the labels in real time by obtaining the number of views of the current idle items, the number of rejections by the querying user, and the number of times the supplementary labels are generated, and comparing the ratio of the maximum number of times the supplementary labels are generated to the number of times the querying user rejects them. When the ratio is greater than the set limit, it means that there may be problems with the current labeling system and an update step needs to be performed to adjust it. This dynamic adjustment strategy based on user feedback helps to ensure the timeliness and accuracy of the labeling system.

[0055] By setting the label update cycle and counting the maximum number of times supplementary labels are generated and the number of times query users are removed within the cycle, the system can further evaluate the necessity and frequency of label updates. When the proportion is greater than or equal to the set limit, the system will dynamically adjust the update cycle according to the proportion to ensure that the label system can promptly reflect market changes and user needs. This cycle adjustment strategy based on statistical data helps to improve the system's response speed and flexibility.

[0056] By implementing the above steps, the system can continuously optimize the label system and improve the accuracy and matching efficiency of labels. This will directly improve the convenience and satisfaction of user queries, because users can find idle items that meet their needs more quickly. At the same time, the system can also dynamically adjust the label update strategy based on user feedback to ensure that the label system is always consistent with user needs, thereby further enhancing the user experience.

[0057] Optionally, after the replacing step, a replacing statistics step is further provided;

[0058] Replacement statistics: count the number of times the idle item tags of the idle items have been replaced, and determine whether the ratio of the number of replacements to the total number of idle item tags associated with the idle items is greater than the set ratio interval. If so, execute the step of moving down in sequence;

[0059] Move down in order: adjust the order of idle item information in the intervention sorting step, move the recommended order of the idle item down, and then execute the push step.

[0060] By adopting the above technical solution, by counting the number of replacements of idle item labels, the system can identify which labels are unstable or inaccurate for specific idle items. This dynamic monitoring helps to timely remove invalid or misleading labels, thereby maintaining the accuracy and timeliness of recommended labels.

[0061] When it is found that the label replacement frequency of an idle item is too high and exceeds the set proportion range, the system will automatically adjust the recommendation order of the item. This adjustment strategy based on user behavior and label stability can ensure that more popular idle items that better meet user needs are displayed first, improving the pertinence and accuracy of recommendations; and frequent label replacement may mean that the label does not match the actual attributes of the idle item, which will cause users to encounter misleading information when browsing. By moving down the order, the system reduces the exposure rate of such misleading information, allowing users to find idle items that meet their needs more quickly.

[0062] Accurate recommendations and rankings can improve user satisfaction and trust in the system. It is easier for users to find their favorite idle items in the system, thereby increasing usage frequency and stickiness.

[0063] The system dynamically adjusts the order of recommendation for idle items according to the label replacement situation, which is actually optimizing the allocation of resources. Allocating limited recommendation resources to idle items with stable labels and high user interest will help improve the overall resource utilization efficiency.

[0064] In the second aspect, the present application provides a management system, which adopts the following technical solution:

[0065] A management system includes the following modules:

[0066] Database: used to store item category information tables, idle item information, and idle item labels;

[0067] Label building module: the input end is connected to the output end of the database, and the output end is connected to the input end of the database, and is used to generate idle item labels according to the item category information table, and store the generated idle item labels in the database;

[0068] Information acquisition module: used to obtain idle item information registered by the publishing user;

[0069] Information processing module: the input end is connected to the output end of the label construction module and the output end of the information acquisition module, and the output end is connected to the input end of the database, and is used to extract the registration keywords according to the idle item information registered by the publishing user, and form the first keyword information, and match the idle item tags according to the first keyword information, and associate the matched idle item tags with the idle items, and store the associated information in the database;

[0070] Information interaction module: the input end is connected to the output end of the information processing module and the output end of the database, and is used to obtain the query requirements of the query user, extract keywords from the query requirements, form second keyword information, match idle item tags according to the second keyword information, and determine idle items according to the idle item tags;

[0071] Push module: the input end is connected to the output end of the information interaction module, and is used to push the idle item information determined according to the idle item tags to the query user;

[0072] Feedback module: The input end is connected to the output end of the information processing module, and is used to feedback the idle items information that has no corresponding demand to the query user;

[0073] Label supplement module: The input end is connected to the output end of the push module, and the output end is connected to the input end of the database. It is used to statistically query the reasons for the rejection of the pushed items rejected by the user, and generate supplementary labels for optimizing the push of the remaining pushed item information according to the supplementary labels, and add the supplementary labels to the idle item labels to identify and associate other idle items with the same attributes.

[0074] By adopting the above technical solution, idle item labels are automatically generated according to item category information, which reduces the burden of manual labeling and improves the efficiency and accuracy of label construction; by extracting keywords and matching idle item labels, automatic classification and association of information are achieved, providing strong support for subsequent queries and push; it can accurately capture the needs of query users and quickly locate relevant idle items through keyword matching. This rapid response capability helps to improve the user's query experience and satisfaction.

[0075] According to the matching results, the idle item information is pushed to the querying user, which realizes the accurate push of information and improves the efficiency of users in obtaining the required information.

[0076] When users query idle items that have no corresponding needs, timely feedback can be given to avoid users' ineffective waiting and enhance the user-friendliness of the system; by statistically querying the reasons for the rejection of the pushed items rejected by users, supplementary labels are generated, and the information of the remaining pushed items is optimized and pushed. This label optimization method based on user feedback can continuously improve the accuracy and effectiveness of labels, making the system more in line with the actual needs of users.

[0077] The addition of supplementary tags not only optimizes the matching effect of the currently pushed items, but also can identify and associate other idle items with the same attributes, further expanding the scope of application and accuracy of the system.

[0078] The entire system has accumulated a large amount of user behavior data and item information through the collaborative work of various modules. These data can be used for subsequent data analysis work, such as user behavior pattern analysis, market trend forecasting, etc., to provide strong support for the continuous optimization and upgrading of the system.

[0079] This technical solution integrates multiple functional modules to achieve efficient management of idle item information, automatic tag construction and matching, accurate push and feedback, and continuous optimization and supplement of tags. These functions work together to improve user experience, improve information management efficiency, promote data mining and analysis, improve push accuracy, and achieve effective management of idle items.

[0080] In summary, the present application includes at least one of the following beneficial technical effects:

[0081] 1. Through the label supplement mechanism, the idle item label system is continuously optimized and improved, so that the push results are closer to the actual needs of users, thereby improving the user experience. In addition, the push data of the post-query users can be optimized according to the feedback of the pre-query users, so that it is more in line with the actual needs of the query users, improving the accuracy of idle item classification management, improving resource reuse efficiency, improving disposal efficiency, and thus promoting the completion of transactions;

[0082] 2. By setting the label update cycle and counting the maximum number of times supplementary labels are generated and the number of times query users are removed within the cycle, the system can further evaluate the necessity and frequency of label updates. When the proportion is greater than or equal to the set limit, the system will dynamically adjust the update cycle according to the proportion to ensure that the label system can reflect market changes and user needs in a timely manner. This cycle adjustment strategy based on statistical data helps to improve the system's response speed and flexibility;

[0083] 3. By integrating multiple functional modules, it realizes the functions of efficient management of idle item information, automatic tag construction and matching, accurate push and feedback, and continuous optimization and supplement of tags. These functions work together to improve user experience, improve information management efficiency, promote data mining and analysis, improve push accuracy, and realize effective management of idle items. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 is a flow chart of the idle items management method in Example 1 of the present application;

[0085] Figure 2 This is a flowchart from the matching calculation step to the sorting step in Example 2 of the present application;

[0086] Figure 3 It is a flowchart from the language library construction step to the semantic expansion step in Example 2 of the present application;

[0087] Figure 4 This is a flowchart from the conflict determination step to the replacement step in Example 2 of the present application;

[0088] Figure 5It is a flowchart from the supplementary judgment step to the period correction step in Example 2 of the present application;

[0089] Figure 6 It is a system diagram of the management system in the embodiment of the present application. DETAILED DESCRIPTION

[0090] The following combination Figures 1 to 6 This application is described in further detail.

[0091] This embodiment discloses a method for managing idle items.

[0092] Example 1: Reference Figure 1 , the idle items management method comprises the following steps:

[0093] Build a database: Establish a database that stores item category information tables, idle item information, and idle item labels;

[0094] Specifically, by building a database, it is convenient to collect statistics on the information of idle items and store the item category information table, so as to generate idle item labels; there is a correlation between the idle item information and the idle item label, that is, there is an index, and the information of idle items belonging to the label can be obtained by locating the idle item label, wherein the idle item label includes category labels, basic labels and supplementary labels; wherein the category label is obtained through the item category information table, and is used to define the category to which the item belongs; the generation of the basic label is determined by information representing the basic characteristics of the item, such as brand, specification, purpose, production date, shelf life, cost performance, high luxury, etc.; the supplementary label is a label that is manually modified or subsequently supplemented by the query user.

[0095] Label construction: Generate idle item labels based on the item category information table, where idle item labels include category labels, basic labels and supplementary labels;

[0096] Information acquisition: obtain the idle items information registered by the publishing user;

[0097] Information processing: including storage segmentation step and the first labeling step;

[0098] Storage segmentation: extract the registered keywords according to the idle items information registered by the publishing user, and form the first keyword information;

[0099] First labeling: matching idle item labels according to the first keyword information, associating the matched idle item labels with the idle items, and storing the associated information in a database;

[0100] Specifically, the idle item information is identified based on the registered idle item information of the publishing user, wherein the idle item information includes instruction manual information, purchase information, usage information, picture information, and other information that can reflect the residual value of the idle items and the basic characteristics of the idle items; by identifying the idle item information, the natural language processing (NLP) technology or a simple string matching algorithm is used to extract the registered keywords and form the first keyword information, and then the idle item tags are matched according to the first keyword information, and the matching idle item tags are associated with the corresponding idle items, and the associated information is stored in the database for easy retrieval of the index.

[0101] Information interaction: including demand acquisition step, demand segmentation step and demand matching step;

[0102] Demand acquisition: obtain query demand information;

[0103] Demand segmentation: extract query keywords based on query demand information to form second keyword information;

[0104] Demand matching: matching idle item tags according to the second keyword information, and determining whether there are idle items that meet the matching tags according to the matched idle item tags combined with the associated information. If so, executing the push step, otherwise, executing the feedback step;

[0105] Push: Push the idle item information to the querying user;

[0106] Feedback: Prompt the querying user that there is no demand for idle items.

[0107] Specifically, the demand information of the querying user is obtained. The demand information is obtained through smart terminals such as mobile phones, computers, tablets and other devices, by accessing relevant web pages, inputting query needs, and forming a link index. After obtaining the query needs, query keywords are extracted according to the query demand information. The query keyword extraction method is the same as the registration keyword extraction method. Second keyword information is formed through the query keyword, and then the second keyword information is matched with the idle item tag. According to the matched idle item tag, the idle items that meet the matching tags are retrieved according to the associated information, and then the idle items that meet all the matching tags are pushed to the terminal of the querying user, and the querying user browses the idle items. If there are no idle items that completely meet the matching tags, the querying user is prompted that there are no idle items that meet the needs.

[0108] Label supplement: including data statistics steps and optimization feedback steps;

[0109] Data statistics: Query the pushed items that are filtered out by users, obtain the reasons for the removal, and generate supplementary labels;

[0110] Optimization feedback: Update the generated supplementary tags to the idle item tags, establish the association information between the supplementary tags and the idle items, and add additional tags to the idle items for subsequent inquiries by other query users.

[0111] Specifically, when a querying user browses the information of idle items, the querying user can obtain the reason for the user's exclusion by setting "not interested", selecting a pop-up label or customizing a label, or generate a supplementary label based on the commonalities of the same products. Other forms of feedback can also be used; the generated supplementary label is then updated to the idle item label, the idle items are supplementarily identified, an association relationship between the supplementary label and the idle items is established, and it is used for label matching by other querying users when they retrieve the information of idle items in the future.

[0112] The implementation principle of this embodiment 1 is as follows: through the construction of labels, various categories of items are distinguished to form idle item labels, and then according to the registered idle item information and corresponding idle item attributes, the idle item labels are associated with the idle items, and the associated information is stored to facilitate the retrieval of idle item information according to the label index; then the query user's needs are obtained, and the query demand information is segmented to form second keyword information, and then the corresponding idle item labels are matched according to the second keyword information, and then all the idle items that meet the requirements are indexed according to the matched idle item labels, and then the idle item information is pushed to the querying user. If there are no idle items that meet all the matched idle item labels, the user is fed back that there are no idle items that meet the requirements and the requirement information needs to be adjusted.

[0113] After the idle item information is pushed to the user, the supplementary tags are generated by obtaining the reasons why the user eliminates the item information that does not meet his needs, and the association between the supplementary tags and the corresponding idle items is established to facilitate the subsequent query index of the user.

[0114] Example 2: Reference Figure 2 , the difference between this embodiment and embodiment 1 is that a matching calculation step and a grading screening step are provided between the demand matching step and the feedback step;

[0115] Matching calculation: Calculate the similarity between the second keyword information and the idle item label, and calculate their sum value to obtain the idle item matching degree;

[0116] Specifically, the similarity between the second keyword information and the idle item label is calculated, wherein the similarity calculation method includes: text similarity calculation based on cosine similarity, string similarity calculation based on edit distance, etc. The similarity value of the category label obtained by calculation is recorded as , the similarity value of the basic label is recorded as , the similarity value of the supplementary label is recorded as Then, according to the number of tags corresponding to the second keyword of the query requirement, the matching degree of idle items is calculated. The calculation model is as follows: , where The number of category tags corresponding to the second keyword information. The value of is 1. , is the number of basic tags corresponding to the second keyword information, , is the number of supplementary tags corresponding to the second keyword information, .

[0117] In other embodiments, the calculation model of the matching degree may introduce weight calculation. ; In the formula is the weight of the category label corresponding to the second keyword information, is the weight corresponding to the basic tag corresponding to the second keyword information, is the weight of the supplementary tag corresponding to the second keyword information, wherein the weight is determined according to the number of idle item tags corresponding to the second keyword information. ,and .

[0118] Grading screening: Determine whether the idle item matching degree is greater than the set push matching value. If so, execute the summary step; otherwise, execute the feedback step;

[0119] Summarize: Summarize the idle items whose matching degree is greater than the set push matching value, and then perform the sorting step;

[0120] Sorting: Sort the idle item information in descending order of matching degree, and then execute the push step.

[0121] Specifically, it is determined whether the matching degree of the idle items is greater than the set push matching value. If so, the idle item information is summarized, and then sorted from large to small according to the matching degree of the idle items, and given to the inquiring user in order. Through sequential recommendation, the push accuracy can be improved, which makes it easier for the inquiring user to quickly find the items he needs.

[0122] The push matching value is Indicates that when the matching degree is calculated by summing all matching values, the matching value is pushed When the matching degree is calculated by summing the product of the weight and the matching value, the matching value is pushed The specific matching value is calculated based on historical data or model training, or is determined through questionnaire surveys or sample user surveys to push matching values.

[0123] In other implementations, the sorting step further includes obtaining the registration time of the idle items, and sorting the idle items with the same matching degree according to the registration time, wherein the sorting is performed in order of the registration time from small to large.

[0124] Specifically, idle items that have not been sold for a long time (i.e. items with a long registration time) will be ranked at the back. The registration time is poor, and the retention time of such idle items in the system may have some problems, such as product problems or price problems. By automatically adjusting the sorting, users are encouraged to update or reprice these items to attract more potential buyers; and when inquiring users browse, the first thing they see is the newly released items, which can bring a sense of freshness and increase the browsing interest and participation of inquiring users; and for inquiring users who frequently visit the system, sorting by registration time helps reduce the possibility of them repeatedly viewing items they have already seen, thereby improving browsing efficiency and browsing coverage, and at the same time making it easier to compare items and obtain the desired idle items faster, further promoting the transaction of idle items and the accuracy of push.

[0125] Reference Figure 3 In other implementations, a language library construction step and a semantic expansion step are also provided after the tag construction step;

[0126] Language library construction: Generate multi-language information based on the item category information table and establish a language library;

[0127] Semantic expansion: Based on the idle item tags and language library, the idle item tags are expanded, and subsidiary tags of the category tags are generated and updated to the idle item tags.

[0128] Specifically, the query users may include foreigners, or people who do not understand Mandarin, or other people who may speak different standard Mandarin. By expanding the idle item tags into multi-language or dialect tags, and then forming subsidiary tags, and updating them to the idle item tags, when the user's query request does not match the corresponding one or the subsidiary tags are used for auxiliary matching, the matching can be more in line with the real needs of the querying user, thereby improving the query accuracy of various users.

[0129] Reference Figure 3 In other implementations, a directional matching step is provided before the demand matching step;

[0130] Directed matching: Based on the query user information, determine the language preference, circle the range of idle item tags in the database, and then perform the demand matching step.

[0131] Specifically, combined with the querying user's information, such as delivery address information, location information, place of origin information or habitual residence information, the language or dialect that the querying user may use is determined through location or habitual residence, and then the relevant language tag range is circled and matched within this range, thereby reducing the different meanings of the same words in different dialects or languages ​​and reducing the inaccuracy of matching. In addition, through the setting of targeted matching, the query speed can be improved.

[0132] Reference Figure 4 In other embodiments, after the optimization feedback step, a conflict determination step is also included;

[0133] Conflict determination: determine whether there is a conflict between the supplementary tag and the existing tag in the association information between the idle items and the idle item tags. If so, execute the tag counting step;

[0134] Tag statistics: Count the number of query users defined as the supplementary tag, and determine whether the number of query users defined as the supplementary tag is greater than the set quantity limit. If so, perform the replacement step;

[0135] Replace: Delete the relationship between the conflicting existing tags and the corresponding idle items.

[0136] Specifically, after generating the supplementary tag, determine whether there is a conflict between the supplementary tag and the existing tag in the association information of the idle item and the idle item tag corresponding to the idle item. If so, count the number of idle items that are defined as the same supplementary tag and the number of conflicts between the supplementary tag and the existing tag, and compare the number of supplementary tags with the set quantity limit value. When it is greater than the set quantity limit value, delete the association information between the conflicting existing tags and the idle item, and retain the association information between the supplementary tag and the idle item; through the replacement relationship, it is explained that the original label of the idle item may be inaccurate or not suitable for the idle item. By querying the user's feedback and correcting it, the idle item label is more in line with the actual information of the idle item, which is convenient for accurate push and promotes the smooth progress of the transaction; the setting of the quantity limit value is determined through historical data simulation or questionnaire survey or data sampling or based on statistical analysis and experimental data.

[0137] Reference Figure 4 In other embodiments, after the data statistics step, an optimization and correction step is also provided;

[0138] Optimization and correction: Perform semantic inversion based on the supplementary tags, generate new second keyword information, and re-execute the demand matching step.

[0139] Specifically, after the querying user eliminates the idle items that he is not interested in, he obtains the reasons for the elimination and generates a new supplementary tag. Then, semantic inversion is performed based on the supplementary tag to form new second keyword information. Then, the demand matching step is re-executed based on the second keyword information to screen the subsequent idle item information again and optimize the promoted information.

[0140] Reference Figure 5 In other implementations, a supplementary judgment step is provided between the data statistics step and the optimization feedback step;

[0141] Supplementary judgment: obtain the browsing times, rejection times and supplementary label generation times of the current idle items, and count the maximum number of times the same supplementary label is generated, and judge whether the maximum number of times the supplementary label is generated accounts for the number of rejection times of the querying user is greater than the set limit ratio. If so, execute the optimization feedback step, otherwise, execute the cycle determination step;

[0142] Cycle determination: Set the label update cycle to , statistical period The maximum number of times of generating supplementary labels and the number of times of query user rejection are determined, and whether the proportion of the maximum number of times of generating supplementary labels in the number of times of query user rejection is greater than or equal to the set limit ratio is determined. If so, the cycle maintenance step is executed, otherwise, the cycle correction step is executed;

[0143] Cycle maintenance: maintain the current tag update cycle and then execute the update step;

[0144] Cycle correction: The cycle is corrected according to the proportion of the maximum number of times that supplementary tags are generated in the number of times that query users are eliminated. The smaller the proportion, the greater the increase in the cycle. As the proportion increases, the trend increases, and then the supplementary judgment step is performed.

[0145] Specifically, the number of views of the current idle item, the number of times the query user has eliminated the item, and the number of times the supplementary label is generated are obtained, the number of each supplementary label generated is counted, and compared to obtain the maximum number of times the same supplementary label is viewed, and then it is determined whether the proportion of the maximum number of times in the number of eliminations is greater than the set limit ratio. If so, the supplementary label is updated to the idle item label, the idle item is marked, and the idle item with the supplementary label is marked; when the proportion of the maximum number of times in the number of eliminations is less than or equal to the set limit ratio, the cycle determination step is executed, and in the cycle Internal statistics again count the number of eliminations and the number of times supplementary labels are generated, and then determine whether the ratio of the maximum number of times supplementary labels are generated to the number of times the query user eliminates is greater than or equal to the set limit ratio. If the condition can be met within the cycle, the optimization feedback step is performed according to the original cycle. If the requirement cannot be met within the period, the period will be adjusted and the statistical time of the period will be extended to make the optimization and update of the supplementary labels more authentic and in line with the actual user feedback. The set limit ratio is determined through historical data simulation, questionnaire survey, data sampling, or based on statistical analysis and experimental data.

[0146] Specifically, the cycle Determination, obtained through data statistics, cycle The initial value is ,Right now , initial value The determination is based on the statistical analysis of historical data; in the cycle correction step, the cycle The calculation model is as follows: ,in, , is the ratio of the maximum number of times supplementary tags are generated to the number of times the query user removes tags, where The value of is determined by the total number of users browsing the idle items. When it is a popular product, The value of is large, and The maximum value of is 1.6, and the preferred value in this embodiment is 1.3.

[0147] Reference Figure 5 In other embodiments, after the replacing step, a replacing statistics step is further provided;

[0148] Replacement statistics: count the number of times the idle item tags of the idle items have been replaced, and determine whether the ratio of the number of replacements to the total number of idle item tags associated with the idle items is greater than the set ratio interval. If so, execute the step of moving down in sequence;

[0149] Move down in order: adjust the order of idle item information in the intervention sorting step, move the recommended order of the idle item down, and then execute the push step.

[0150] Specifically, when there are multiple label replacements in the association information between an idle item and the idle item label, there may be a problem with the initial label definition of the idle item. Through comparison, the idle items that may have problems are marked. When sorting, the order of the marked item information is lowered, which increases the user's browsing of idle items with more real labels, thereby improving the browsing quality of the querying users. The set proportion interval is determined based on the statistical analysis of historical data.

[0151] The implementation principle of Example 2 of the present application is as follows: the idle item label is expanded through semantic expansion so that it can be applicable to multi-language environment label indexing. After obtaining the second keyword information, the area is defined within the expanded subsidiary label through targeted matching, and then the matching degree is calculated, and graded screening is performed. The idle item information that meets the conditions is counted, and then sorting intervention is performed according to the matching degree, registration time intervention, and replacement statistics. Then, the idle item information is pushed to the querying user in order.

[0152] And in the process of query users browsing, statistics are conducted to find that for the same idle item, most of the query users approve the supplementary tag. If so, the supplementary tag is associated with the idle item. Otherwise, the cycle is determined to determine whether the ratio of the maximum number of times the supplementary tag is generated to the number of times the query users reject the item is greater than the set limit ratio within the set cycle. If it can be met, the current update cycle is maintained. Otherwise, the cycle is adjusted upward through the cycle calculation model to optimize the supplementary tag update time to make it more in line with reality.

[0153] After the supplementary tag is associated with the idle item, determine whether there is a conflict between the supplementary tag and other existing tags of the same idle item. When there is a conflict, determine whether the existing tag is incorrectly identified by counting whether the supplementary tag is recognized by most querying users, and then delete and replace it. In the process of generating the supplementary tag, the second keyword information of the current query user is updated through optimization and correction steps, and subsequent push is optimized.

[0154] The embodiment of the present application also discloses a management system.

[0155] Reference Figure 6 , management system, including the following modules:

[0156] Database: used to store item category information tables, idle item information, and idle item labels;

[0157] Label construction module: the input end is connected to the output end of the database, and the output end is connected to the input end of the database, and is used to generate idle item labels according to the item category information table, and store the generated idle item labels in the database, and generate multiple language information according to the item category information table, and establish a language library, expand the idle item labels to form subsidiary labels, and update the subsidiary labels to the idle item labels;

[0158] Information acquisition module: used to obtain idle item information registered by the publishing user;

[0159] Information processing module: the input end is connected to the output end of the label construction module and the output end of the information acquisition module, and the output end is connected to the input end of the database, and is used to extract the registration keywords according to the idle item information registered by the publishing user, and form the first keyword information, and match the idle item tags according to the first keyword information, and associate the matched idle item tags with the idle items, and store the associated information in the database;

[0160] Information interaction module: The input end is connected to the output end of the information processing module and the output end of the database, and is used to obtain the query requirements of the query user, and extract the query requirements by keywords to form second keyword information, and then circle the idle item tag range in the database according to the second keyword information and the query user information, and then match the idle item tags, and determine the idle items according to the idle item tags, and calculate the matching degree, and then sort the idle item information according to the matching degree, registration time and historical replacement statistical data;

[0161] Push module: The input end is connected to the output end of the information interaction module, and is used to determine the idle item information according to the idle item tags, and push it to the query user according to the sorting;

[0162] Feedback module: The input end is connected to the output end of the information processing module, and is used to feedback the idle items information that has no corresponding demand to the query user;

[0163] Label supplement module: the input end is connected to the output end of the push module, and the output end is connected to the input end of the database, which is used to statistically query the reasons for the rejection of the pushed items rejected by the user, and generate supplementary labels, which are used to optimize the push of the remaining pushed item information according to the supplementary labels, and statistically determine whether the supplementary labels are labels recognized by multiple users. If so, the supplementary labels are added to the idle item labels, and other idle items with the same attributes are identified and associated, and the semantics of the supplementary labels are reversed to form the second keyword information; otherwise, the semantics of the supplementary labels are reversed to form the second keyword information, and then the cycle is determined to determine whether the replacement step of the supplementary labels is completed within the set cycle. If not, the cycle is corrected and the cycle is extended.

[0164] This system processes the idle item information and associates the idle item tags through the tag construction module, information acquisition module, information processing module, information interaction module, push module, feedback module and tag supplement module, then segment the query user demand information to form the second keyword information, then calculate the matching degree between the second keyword information and the idle item tag, and screen it, then sort the idle item information according to the registration time and the feedback of the historical replacement statistics, and then push it, and in the process of users screening and eliminating the idle item information, through data statistics, supplement the idle item tags or replace the existing tags, optimize the accuracy of the idle item information, and feedback to the subsequent idle item push, so that the push results are more accurate and the transaction is promoted.

[0165] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for managing idle items, characterized in that: The following steps are involved: Build a database: Establish a database that stores item category information tables, idle item information, and idle item labels; Label construction: Generate idle item labels based on the item category information table, where idle item labels include category labels, basic labels and supplementary labels; Information acquisition: obtain the idle items information registered by the publishing user; Information processing: including storage segmentation step and the first labeling step; Storage segmentation: extract the registered keywords according to the idle items information registered by the publishing user, and form the first keyword information; First labeling: matching idle item labels according to the first keyword information, associating the matched idle item labels with the idle items, and storing the associated information in a database; Information interaction: including demand acquisition step, demand segmentation step and demand matching step; Demand acquisition: obtain query demand information; Demand segmentation: extract query keywords based on query demand information to form second keyword information; Demand matching: matching idle item tags according to the second keyword information, and determining whether there are idle items that meet the matching tags according to the matched idle item tags combined with the associated information. If so, executing the push step, otherwise, executing the feedback step; Push: Push the idle item information to the querying user; Feedback: Prompt the querying user that there is no idle item in demand; Label supplement: including data statistics steps and optimization feedback steps; Data statistics: Query the pushed items that are filtered out by users, obtain the reasons for the removal, and generate supplementary labels; Optimization feedback: Update the generated supplementary tags to the idle item tags, establish the association information between the supplementary tags and the idle items, and add supplementary tags to the idle items for subsequent inquiries by other query users; A supplementary judgment step is also provided between the data statistics step and the optimization feedback step; Supplementary judgment: obtain the browsing times, rejection times and supplementary label generation times of the current idle items, and count the maximum number of times the same supplementary label is generated, and judge whether the maximum number of times the supplementary label is generated accounts for the number of rejection times of the querying user is greater than the set limit ratio. If so, execute the optimization feedback step, otherwise, execute the cycle determination step; Cycle determination: Set the label update cycle to , statistical period The maximum number of times of generating supplementary labels and the number of times of query user rejection are determined, and whether the proportion of the maximum number of times of generating supplementary labels in the number of times of query user rejection is greater than or equal to the set limit ratio is determined. If so, the cycle maintenance step is executed, otherwise, the cycle correction step is executed; Cycle maintenance: maintain the current tag update cycle and then execute the update step; Cycle correction: The cycle is corrected according to the proportion of the maximum number of times that supplementary tags are generated in the number of times that query users are eliminated. The smaller the proportion, the greater the increase in the cycle. As the proportion increases, the trend increases, and then the supplementary judgment step is performed.

2. The idle items management method according to claim 1, characterized in that: A matching calculation step and a grading and screening step are also provided between the demand matching step and the feedback step; Matching calculation: Calculate the similarity between the second keyword information and the idle item label, and calculate their sum value to obtain the idle item matching degree; Grading and screening: determine whether the matching degree of idle items is greater than the set push matching value. If so, execute the summary step; Otherwise, the feedback step is executed; Summarize: Summarize the idle items whose matching degree is greater than the set push matching value, and then perform the sorting step; Sorting: Sort the idle item information in descending order of matching degree, and then execute the push step.

3. The idle items management method according to claim 2, characterized in that: The sorting step further includes obtaining the registration time of the idle items, and sorting the idle items with the same matching degree according to the registration time, wherein the sorting is performed in order of the registration time from small to large.

4. The idle items management method according to claim 2, characterized in that: After the tag building step, there are also a language library building step and a semantic expansion step; Language library construction: Generate multi-language information based on the item category information table and establish a language library; Semantic expansion: Based on the idle item tags and language library, the idle item tags are expanded, and subsidiary tags of the category tags are generated and updated to the idle item tags.

5. The idle items management method according to claim 4, characterized in that: A directional matching step is also provided before the demand matching step; Directed matching: Based on the query user information, determine the language preference, circle the range of idle item tags in the database, and then perform the demand matching step.

6. The idle items management method according to any one of claims 2 to 5, characterized in that: After the optimization feedback step, a conflict determination step is also included; Conflict determination: determine whether there is a conflict between the supplementary tag and the existing tag in the association information between the idle items and the idle item tags. If so, execute the tag counting step; Tag statistics: Count the number of query users defined as the supplementary tag, and determine whether the number of query users defined as the supplementary tag is greater than the set quantity limit. If so, perform the replacement step; Replace: Delete the relationship between the conflicting existing tags and the corresponding idle items.

7. The idle items management method according to claim 6, characterized in that: After the data statistics step, an optimization and correction step is also provided; Optimization and correction: Perform semantic inversion based on the supplementary tags, generate new second keyword information, and re-execute the demand matching step.

8. The idle items management method according to claim 6, characterized in that: After the replacement step, a replacement statistics step is also provided; Replacement statistics: count the number of times the idle item tags of the idle items have been replaced, and determine whether the ratio of the number of replacements to the total number of idle item tags associated with the idle items is greater than the set ratio interval. If so, execute the step of moving down in sequence; Move down in order: adjust the order of idle item information in the intervention sorting step, move the recommended order of the idle item down, and then execute the push step.

9. A management system for implementing the idle items management method according to any one of claims 1 to 8, characterized in that: Includes the following modules: Database: used to store item category information tables, idle item information, and idle item labels; Label building module: the input end is connected to the output end of the database, and the output end is connected to the input end of the database, and is used to generate idle item labels according to the item category information table, and store the generated idle item labels in the database; Information acquisition module: used to obtain idle item information registered by the publishing user; Information processing module: the input end is connected to the output end of the label construction module and the output end of the information acquisition module, and the output end is connected to the input end of the database, and is used to extract the registration keywords according to the idle item information registered by the publishing user, and form the first keyword information, and match the idle item tags according to the first keyword information, and associate the matched idle item tags with the idle items, and store the associated information in the database; Information interaction module: the input end is connected to the output end of the information processing module and the output end of the database, and is used to obtain the query requirements of the query user, extract keywords from the query requirements, form second keyword information, match idle item tags according to the second keyword information, and determine idle items according to the idle item tags; Push module: the input end is connected to the output end of the information interaction module, and is used to push the idle item information determined according to the idle item tags to the query user; Feedback module: The input end is connected to the output end of the information processing module, and is used to feedback the idle items information that has no corresponding demand to the query user; Label supplement module: The input end is connected to the output end of the push module, and the output end is connected to the input end of the database. It is used to statistically query the reasons for the rejection of the pushed items rejected by the user, and generate supplementary labels for optimizing the push of the remaining pushed item information according to the supplementary labels, and add the supplementary labels to the idle item labels to identify and associate other idle items with the same attributes.

Images