Search

CN-121981804-A - Commodity shopping guide method and device, equipment and medium thereof

CN121981804ACN 121981804 ACN121981804 ACN 121981804ACN-121981804-A

Abstract

Responding to shopping guide dialogue events of a user and customer service of an independent station store, and acquiring an initial text describing the current purchase intention of the user; when the fact that the specific degree of the current purchasing intention is not up to standard is detected, a preference feature data set matched with the initial text is determined based on the interactive behavior sequence of the user, a plurality of shopping guide recommended commodities matched with the initial text are recalled, preference priority ordering is conducted on the shopping guide recommended commodities based on the preference feature data set, a commodity recommendation list is obtained, and the commodity recommendation list is pushed to the user to display the shopping guide recommended commodities in the list to the user. The application provides customer service for commodity shopping guide for the user, can greatly improve the shopping experience of the user and promote the achievement of commodity transaction.

Inventors

  • XIE YIHUA

Assignees

  • 广州商耘网络科技有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (10)

  1. 1. The commodity shopping guide method is characterized by comprising the following steps: Responding to shopping guide dialogue events of the user and customer service of the independent station store, and acquiring an initial text describing the current purchase intention of the user; when detecting that the specific degree of the initial text describing the current buying intention does not reach the standard, determining a preference characteristic data set matched with the initial text based on the interactive behavior sequence of the user; Recalling a plurality of shopping guide recommended commodities matched with the initial text, and carrying out preference priority ordering on the shopping guide recommended commodities based on the preference characteristic data set to obtain a commodity recommendation list; Pushing the commodity recommendation list to the user so as to display shopping guide recommended commodities in the list to the user.
  2. 2. The merchandise shopping guide method of claim 1, wherein determining a preference feature data set that matches the initial text based on the interactive behavior sequence of the user comprises the steps of: the method comprises the steps of obtaining an interactive behavior sequence of a user, wherein the interactive behavior sequence comprises an interactive behavior type, interactive commodity information, an interactive triggering time stamp and an interactive completion time length which correspond to each time of interaction of the user at an independent station shop, and the interactive commodity information comprises commodity categories, commodity description keywords and commodity pictures of interactive commodities; Determining the recall commodity category matched with the initial text, identifying all shopping preference subsequences and real-time performance tags thereof associated with the commodity category in the interactive behavior sequence, and determining preference characteristic data corresponding to the shopping preference subsequences based on the real-time performance tags, wherein the preference characteristic data is used for forming a preference characteristic data set matched with the initial text.
  3. 3. The merchandise shopping guide method according to claim 2, wherein identifying all shopping preference sub-sequences and real-time performance tags thereof associated with the category of merchandise in the interactive behavior sequence and determining preference feature data corresponding to the shopping preference sub-sequences thereof based on the real-time performance tags comprises the steps of: Adopting a preset recognition prompt text to guide a large language model, and determining all shopping preference subsequences and real-time performance tags thereof in the interactive behavior sequence, wherein the shopping preference subsequences are associated with the recalled commodity category, and the real-time performance tags comprise preference behavior types and preference event time stamps; Adopting a preset feature extraction strategy corresponding to the preference behavior type, and extracting corresponding commodity feature information according to the interactive commodity information in the shopping preference subsequence associated with the tag; acquiring a basic performance score and a time attenuation coefficient which correspond to the preference behavior type, and performing value attenuation processing on the basic performance score based on the preference event time stamp and the time attenuation coefficient to acquire a preference performance score of the commodity characteristic information; and combining the commodity characteristic information and the preference performance score thereof into preference characteristic data.
  4. 4. The merchandise shopping guide method according to claim 1, wherein a plurality of shopping guide recommended merchandise items matched with the initial text are recalled, and the shopping guide recommended merchandise items are prioritized based on the preference feature data set, and a merchandise recommendation list is obtained, comprising the steps of: determining a purchase demand text corresponding to the initial text by adopting a preset intention analysis model, and recalling a plurality of shopping guide recommended commodities in the independent station shops, which are matched with the purchase demand text; And determining the correlation between the commodity characteristic information in each preference characteristic data set and the commodity information of the shopping guide recommended commodity, and correspondingly weighting and summing the preference performance scores of each commodity characteristic information to obtain the preference recommendation score of the shopping guide recommended commodity.
  5. 5. The merchandise shopping guide method of claim 1, wherein pushing the merchandise recommendation list to the user to present the shopping guide recommended merchandise in the list to the user comprises the steps of: Responding to a shopping guide commodity display event, and displaying a dialogue component corresponding to the shopping guide dialogue event in a preset shopping guide page in a thumbnail manner; and displaying the shopping guide recommended commodities in the commodity recommendation list correspondingly in the shopping guide page according to a preset commodity display layout strategy.
  6. 6. The merchandise shopping guide method according to claim 1, wherein when it is detected that the initial text describes that the specific degree of the current purchase intention does not reach the standard, before determining the preference feature data set matching the initial text based on the interactive behavior sequence of the user, comprising the steps of: collecting a plurality of historical initial texts as samples respectively, describing whether the specific degree of the corresponding buying intention meets the standard according to each sample, and labeling a corresponding supervision label; All samples are associated with their supervision labels to form a dataset, and the intent expression model is trained to a convergence state by adopting the dataset for detecting the initial text.
  7. 7. The merchandise purchase-guiding method of claim 1, wherein prior to responding to the purchase-guiding dialogue event of the user with the customer service of the independent station store, comprising: And responding to the shopping guide page display event, loading the shopping guide page to display an advertisement component and a dialogue component therein, wherein the advertisement component displays at least one advertisement commodity related to the user in the independent station store, and the dialogue component is used for conducting shopping guide dialogue between the user and the independent station store customer service.
  8. 8. A merchandise shopping guide device, comprising: The event response module is used for responding to shopping guide dialogue events of the user and customer service of the independent station store and acquiring an initial text describing the current purchase intention of the user; The preference supplementing module is used for determining a preference characteristic data set matched with the initial text based on the interactive behavior sequence of the user when the fact that the specific degree of the initial text describing the current purchase intention does not reach the standard is detected; The list construction module is used for recalling a plurality of shopping guide recommended commodities matched with the initial text, and carrying out preference priority ordering on the shopping guide recommended commodities based on the preference characteristic data set to obtain a commodity recommendation list; And the recommendation display module is used for pushing the commodity recommendation list to the user so as to display shopping guide recommended commodities in the list to the user.
  9. 9. A computer device comprising a central processor and a memory, characterized in that the central processor is arranged to invoke a computer program stored in the memory for performing the steps of the method according to any of claims 1 to 7.
  10. 10. A computer-readable storage medium, characterized in that it stores in the form of computer-readable instructions a computer program implemented according to the method of any one of claims 1 to 7, which, when invoked by a computer, performs the steps comprised by the corresponding method.

Description

Commodity shopping guide method and device, equipment and medium thereof Technical Field The present application relates to the technical field of electronic commerce, and in particular, to a commodity shopping guide method, and a corresponding apparatus, computer device, and computer readable storage medium thereof. Background In existing online shopping environments, users typically find desired goods by using an e-commerce search service. However, many times users often have only one ambiguous or generalized shopping idea, or the shopping idea is difficult to express, resulting in the fact that their entered query text may lack specific details. This ambiguous intent representation makes it difficult to directly return accurate merchandise results, and the user has to step through multiple attempts at different search terms, manually screen and view large amounts of irrelevant or less relevant merchandise information to make explicit demands. The whole process is low in efficiency and long in time consumption, and a user is easy to feel confused or lose patience in repeated attempts, so that the final purchase intention and commodity yield are reduced. Meanwhile, while most provide customer service dialogue functions for users, conventional online customer service systems are primarily targeted to pre-sale consultation and after-sale services. A typical application scenario is that a user has generated interest in a specific commodity or a certain class of specific commodities, and then inquires about specific problems such as specifications, quality, logistics or sales promotion policies of the commodity. In other words, initiation of a traditional customer service interaction is often premised on a user having a degree of explicit purchase intent. Therefore, when a user wishes to enjoy a commodity shopping guide service provided by a sales person who makes a same online shopping, a conventional online customer service cannot provide the service. Therefore, the application develops another way to solve the problem to be solved in the prior art. Disclosure of Invention It is therefore a primary object of the present application to solve at least one of the above problems and provide a commodity shopping guide method, and corresponding apparatus, computer device, and computer-readable storage medium. In order to meet the purposes of the application, the application adopts the following technical scheme: The commodity shopping guide method provided by the application, which is suitable for one of the purposes of the application, comprises the following steps: responding to a commodity shopping guide event triggered by a user, acquiring a commodity description text of a shopping guide commodity corresponding to the event, and inquiring whether a preset user portrait library stores a portrait tag set of the user or not, wherein the portrait tag set is obtained based on consumption preference reasoning on long-term historical interaction behaviors of the user; When the user portrait library stores portrait tag sets of the user, determining the correlation degree between each portrait tag in the portrait tag sets and the commodity description text, and screening portrait tag structures with the correlation degree exceeding a preset threshold to generate dependency information; based on the generated dependency information and the target prompt text corresponding to the commodity description text structure, the target prompt text is used for guiding a large language model to convert the commodity description text into personalized marketing text of the consumer preference; and displaying personalized marketing texts of the shopping guide commodities to the user according to a preset shopping guide marketing strategy. On the other hand, the commodity shopping guide device provided by the application is suitable for one of the purposes of the application, and comprises an event response module, a preference supplementing module, a list construction module and a recommendation display module, wherein the event response module is used for responding to shopping guide dialogue events of users and independent station shops and acquiring initial texts describing current purchase intentions of the users, the preference supplementing module is used for determining a preference characteristic data set matched with the initial texts based on an interaction behavior sequence of the users when detecting that the specific degree of the current purchase intentions described by the initial texts does not reach standards, the list construction module is used for recalling a plurality of shopping guide recommended commodities matched with the initial texts and carrying out preference priority ordering on the shopping guide recommended commodities based on the preference characteristic data set to obtain a commodity recommendation list, and the recommendation display module is used for pushing the commodity recommendation list to