CN116611897B - Message reminding method and system based on artificial intelligence - Google Patents
Message reminding method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN116611897B CN116611897B CN202310884919.0A CN202310884919A CN116611897B CN 116611897 B CN116611897 B CN 116611897B CN 202310884919 A CN202310884919 A CN 202310884919A CN 116611897 B CN116611897 B CN 116611897B
- Authority
- CN
- China
- Prior art keywords
- commodity
- sharing
- information
- shared
- message
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Recommending goods or services
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/042—Knowledge-based neural networks; Logical representations of neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/18—Status alarms
- G08B21/24—Reminder alarms, e.g. anti-loss alarms
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Accounting & Taxation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Emergency Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a message reminding method and a system based on artificial intelligence, and relates to the technical field of message reminding; determining the detail page description matching degree by using a detail page description matching model based on information of the commodity detail pages of the shared commodity in each commodity sharing message; determining the matching degree of the sharing description by using a sharing description matching model based on information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message; constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes; processing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the graph neural network model to determine a target commodity node; the method can screen and remind the high-quality commodity information meeting the actual demands of the users from the shopping group.
Description
Technical Field
The invention relates to the technical field of message reminding methods, in particular to a message reminding method and system based on artificial intelligence.
Background
With the popularity of electronic commerce and social networks, more and more users join various shopping groups through social software to obtain shared commodity information of other users in the shopping group, and communicate and interact with other users. However, due to the reasons of numerous and complicated shared information, different tastes of users, poor information quality and the like, the shared commodity information can frequently remind the users to cause bad experience to the users, commodity quality of shopping groups is not guaranteed, and the users also need to spend a great deal of time to select commodities. In the prior art, most of the methods for screening and reminding commodity information in a shopping group are to construct a keyword list by text analysis and keyword extraction of commodity information in the shopping group, and then match keywords with the purchasing demands of users so as to screen commodities meeting the demands of the users.
Therefore, how to screen and remind the information of the good quality commodity meeting the actual demands of the user from the shopping group is a current urgent problem to be solved.
Disclosure of Invention
The invention mainly solves the technical problem of how to screen and remind the information of the high-quality commodity meeting the actual demands of users from shopping clusters.
According to a first aspect, the present invention provides an artificial intelligence based message reminding method, comprising: acquiring a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared characters corresponding to the shared commodity; acquiring information of a sharing user corresponding to each commodity sharing message and information of a commodity detail page of the shared commodity in each commodity sharing message; calculating user similarity based on the information of the sharing users corresponding to each commodity sharing message and the information of the shared users; determining the detail page description matching degree by using a detail page description matching model based on the information of the commodity detail pages of the shared commodity in each commodity sharing message; determining the matching degree of the sharing description by using a sharing description matching model based on information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message; constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales and scoring information; processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model to determine a target commodity node, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node; and carrying out message reminding on the shared user based on the commodity sharing message corresponding to the target commodity node.
Further, the calculating the user similarity based on the information of the sharing user corresponding to each commodity sharing message and the information of the shared user includes: and calculating the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user, and calculating the similarity of the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user according to the Hamming distance.
Further, the method for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node may include vibration reminding, sound reminding and window reminding.
Further, the information of the commodity detail page of the shared commodity in each commodity sharing message comprises video introduction information, picture introduction information and text introduction information.
Still further, the method further comprises: and if the matching degree of the sharing description of the commodity sharing message is smaller than the matching threshold of the sharing description, the popup window reminds the shared user to carefully identify the commodity sharing message smaller than the matching threshold of the sharing description.
According to a second aspect, the present invention provides an artificial intelligence based message alert system comprising: the first acquisition module is used for acquiring a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared words corresponding to the shared commodity; the second acquisition module is used for acquiring information of the sharing users corresponding to each commodity sharing message and information of commodity detail pages of the shared commodities in each commodity sharing message; the user similarity calculation module is used for calculating user similarity based on the information of the sharing users corresponding to each commodity sharing message and the information of the shared users; the detail page matching module is used for determining the detail page description matching degree by using a detail page description matching model based on the information of the commodity detail pages of the shared commodities in each commodity sharing message; the sharing description matching module is used for determining the matching degree of the sharing description based on the information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message by using the sharing description matching model; the construction module is used for constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales and scoring information; the target commodity node determining module is used for processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node; and the reminding module is used for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node.
Further, the user similarity calculation module is further configured to calculate a SimHash value of the information of the shared user corresponding to each commodity sharing message and a SimHash value of the information of the shared user, and calculate a similarity between the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user according to a hamming distance.
Further, the method for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node may include vibration reminding, sound reminding and window reminding.
Further, the information of the commodity detail page of the shared commodity in each commodity sharing message comprises video introduction information, picture introduction information and text introduction information.
Still further, the system is further configured to: and if the matching degree of the sharing description of the commodity sharing message is smaller than the matching threshold of the sharing description, the popup window reminds the shared user to carefully identify the commodity sharing message smaller than the matching threshold of the sharing description.
The invention provides a message reminding method and a message reminding system based on artificial intelligence, wherein the method comprises the steps of obtaining a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared words corresponding to the shared commodity; acquiring information of a sharing user corresponding to each commodity sharing message and information of a commodity detail page of the shared commodity in each commodity sharing message; calculating user similarity based on the information of the sharing users corresponding to each commodity sharing message and the information of the shared users; determining the detail page description matching degree by using a detail page description matching model based on the information of the commodity detail pages of the shared commodity in each commodity sharing message; determining the matching degree of the sharing description by using a sharing description matching model based on information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message; constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales and scoring information; processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model to determine a target commodity node, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node; and carrying out message reminding on the shared user based on the commodity sharing message corresponding to the target commodity node, wherein the method can screen and remind the high-quality commodity information meeting the actual demands of the user from the shopping group.
Drawings
FIG. 1 is a schematic flow chart of an artificial intelligence based message reminding method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a message reminding system based on artificial intelligence according to an embodiment of the present invention.
Detailed Description
In the embodiment of the invention, an artificial intelligence based message reminding method is provided as shown in fig. 1, and the artificial intelligence based message reminding method comprises the following steps S1-S8:
step S1, acquiring a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared words corresponding to the shared commodities.
Shopping group refers to a group formed by one or more users in the application of an e-commerce platform, social media and the like, and the users can release relevant contents such as commodity information, evaluation, sharing and the like in the group. For example, a "women's share" shopping group on Taobao may consist of some users who like to purchase women's, and the users may share the contents of shopping hearts, wearing skills, preferential information, etc. in the group.
The commodity sharing message refers to a message which is shared by the user in the shopping group and contains commodity information and text description. Each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared characters corresponding to the shared commodity.
The shared commodity information refers to information of the shared commodity, for example, the shared commodity information includes detailed information such as commodity price, commodity type, commodity name, sales volume, scoring information and the like. For another example, the merchandise information of a "dress" may include information about the brand, material, size, and application of the dress. The sharing text refers to text description or evaluation information of the commodity in the shopping group by the user. For example: the user's small piece shares a piece of shared commodity information containing' one-piece dress 'in the shopping group, and is matched with' the skirt is hundred percent high-quality long staple cotton, five brand manufacturers produce in the whole country and send the brand scarves-! "share text.
The information of the shared user is information of the user who receives a plurality of commodity sharing messages in the shopping group. The information of the shared users comprises user names, sexes, ages, regions, shopping class preference, historical purchase amount, historical evaluation information, historical purchase orders and the like.
The user name represents the user name when the user registers on the e-commerce platform or social media. For example, the Taobao user name is "alice123" and the micro-credit user name is "jacky456". Gender refers to the gender information of the user, and can be used for accurate marketing in certain scenes. Such as male, female, or privacy, etc. Age refers to age information of a user, and can be used for accurate marketing in certain scenes. Such as 18 years, 25 years, 35 years, etc. The region refers to the geographical position information of the user, and can be used for regional accurate marketing. Such as Shenzhen city, guangdong province, nanjing city, jiangsu province, new York City, U.S. Shopping category preference refers to the preference degree of users for different commodity categories, and can be used for personalized recommendation and accurate marketing. Such as, for example, to like to purchase women's clothing, digital products, food, etc. The historical purchase amount refers to the historical purchase amount of the user on the e-commerce platform or social media, and may provide an auxiliary decision. For example, the user's historical purchase amount on a certain e-commerce platform is 1000 yuan. The historical evaluation information refers to the historical evaluation information of the user on an e-commerce platform or social media, and can provide auxiliary decisions. For example, a user may evaluate a certain cell phone as "good quality, worth purchasing". Historical purchase orders refer to historical purchase order information of users on an e-commerce platform or social media, and can provide auxiliary decisions for business personnel. For example, the historical purchase order information of the user on a certain e-commerce platform includes purchase time, commodity name, price, etc.
In some embodiments, the message data in the shopping group can be obtained through a crawler technology, and commodity information and shared text in the message data are extracted. Specifically, message data of shopping groups in an electronic commerce platform or social media can be crawled through an API interface or a third-party tool, then the message is analyzed and processed by using natural language processing and data mining technology, commodity information and shared words are extracted from the message data, and the commodity information and the shared words are stored in a database for later use.
And S2, acquiring information of the sharing users corresponding to each commodity sharing message and information of commodity detail pages of the shared commodities in each commodity sharing message.
And the information of the sharing user corresponding to each commodity sharing message is the information of the user actively sharing the message. For example, a human sharing user actively sending out a commodity sharing message. The information of the sharing users corresponding to each commodity sharing message comprises a user name, gender, age, region, shopping class preference, historical purchase amount, historical evaluation information, historical purchase orders and the like.
The commodity detail page information refers to related information of commodity detail pages in an electronic commerce platform or social media, and the commodity detail page information can comprise video introduction information, picture introduction information and text introduction information. Video presentation information refers to video presentation information provided in item detail pages on an electronic commerce platform or social media. For example, an air purifier in an electronic mall of a certain household provides video of the functional introduction in a detail page of the product. The picture introduction information refers to picture introduction information provided in the commodity detail page on the e-commerce platform or social media. For example, a plurality of commodity pictures in a Taobao commodity detail page show commodity appearance, detail and function introduction under different angles and scenes. The text introduction information refers to text introduction information provided in the commodity detail page on the electronic commerce platform or the social media, and comprises characteristics, brands, parameters, using methods and the like of commodities. For example, a certain milk powder commodity provides detailed text description in a commodity detail page on the Taobao, including components, efficacy, place of origin, applicable age, etc.
And S3, calculating the user similarity based on the information of the shared user corresponding to each commodity sharing message and the information of the shared user.
In some embodiments, the similarity between the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user can be calculated by calculating the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user through a hamming distance, and the similarity is taken as the user similarity.
In some embodiments, the steps of calculating the SimHash value of the information of the shared user and calculating the SimHash value of the information of the shared user may include the steps of word segmentation, hash calculation, weighting, merging, dimension reduction, and the like. For example, word segmentation: firstly, performing word segmentation on the SimHash value for calculating the information of the shared user and the SimHash value for calculating the information of the shared user, and extracting feature vectors. And setting weight (weight) for the feature vector; hash calculation: calculating a hash value of each feature vector through a hash function, wherein the hash value is an n-bit signature consisting of binary numbers 01; weighting: on the basis of the hash value, weighting all feature vectors, namely WHAsh weight, wherein when the weight is 1, the hash value and the weight are multiplied positively, and when the weight is 0, the hash value and the weight are multiplied negatively; combining: accumulating the weighted results of the feature vectors to form a serial string; dimension reduction: and setting 1 if the accumulated result is greater than 0, otherwise setting 0, so as to obtain a SimHash value for calculating the information of the shared user and a SimHash value for calculating the information of the shared user, and then calculating the similarity between the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user according to the Hamming distance, wherein the similarity is used as the user similarity.
In some embodiments, user similarity may also be determined by a user similarity model, which is a deep neural network model that includes a deep neural network (Deep Neural Networks, DNN), which is one implementation of artificial intelligence. The input of the user similarity model is the information of the sharing user corresponding to each commodity sharing message and the information of the shared user, and the output of the user similarity model is the user similarity.
The user similarity can be used for assisting commodity information recommendation, for example, the higher the user similarity between the sharing user and the shared user is, the closer shopping habits of the two users are, and the commodity recommended by the sharing user can be recommended to the shared user preferentially.
And S4, determining the detail page description matching degree by using a detail page description matching model based on the information of the detail pages of the shared commodities in each commodity sharing message.
Because the information of the commodity detail page comprises video introduction information, picture introduction information and text introduction information, the information of the commodity detail page needs to be processed to determine whether a plurality of introduction information in the commodity detail page is matched, namely whether the video introduction information, the picture introduction information and the text introduction information are consistent, for example, if the commodity is a mobile phone, the video introduction information displays 8G of the running memory of the mobile phone, 4G of the running memory of the mobile phone is displayed in the picture introduction information, 2G of the running memory of the mobile phone is displayed in the text introduction information, and therefore, the serious discrepancy of the various introduction information in the commodity detail page is indicated, and false propaganda may exist for the commodity, and the commodity information is not suitable for being recommended to users.
The detail page description matching model is a long-short-term neural network model. The Long-Short Term neural network model includes a Long-Short Term neural network (LSTM). The long-term and short-term neural network model can process sequence data with any length, capture sequence information and output results based on the association relationship of front data and rear data in the sequence. The detail page description matching model comprehensively considers video introduction information, picture introduction information and text introduction information of each time point, and finally determines the detail page description matching degree. The detail page description matching model can be obtained by training a training sample through a gradient descent method.
The detail page description matching degree indicates the matching degree of each item of information in the commodity detail page output through the detail page description matching model. The detail page description matching degree can be a numerical value between 0 and 1, and the larger the numerical value is, the higher the matching degree is, and the more suitable the commodity is for being recommended to shared users. For example, if the commodity is a dress, and the materials of the dress are mercerized cotton in the video introduction information, the picture introduction information and the text introduction information, the detailed page description is consistent, and the matching degree of the detailed page description output by the detailed page description matching model can be 0.98. For another example, if the commodity is a dress, and the materials of the dress displayed in the video introduction information, the picture introduction information and the text introduction information are mercerized cotton, combed cotton and knitted plain cloth respectively, the detailed page description is not consistent, and the matching degree of the detailed page description output by the detailed page description matching model can be 0.02.
The input of the detail page description matching model is video introduction information, picture introduction information and text introduction information in the commodity detail page of the shared commodity, and the output of the detail page description matching model is detail page description matching degree.
In some embodiments, the detail page description matching model comprises a text-picture matching sub-model, a text-video matching sub-model, a picture-video matching sub-model and a comprehensive matching degree output sub-model, wherein the input of the text-picture matching sub-model is text introduction information and picture introduction information, and the output of the text-picture matching sub-model is the matching degree of the text introduction information and the picture introduction information. The input of the text video matching sub-model is text introduction information and video introduction information, and the output of the text video matching sub-model is the matching degree of the text introduction information and the video introduction information. The input of the picture video matching sub-model is picture introduction information and video introduction information, and the output of the picture video matching sub-model is the matching degree of the picture introduction information and the video introduction information. The input of the comprehensive matching degree output submodel is the matching degree of the text introduction information and the picture introduction information, the matching degree of the text introduction information and the video introduction information and the matching degree of the picture introduction information and the video introduction information, and the output of the comprehensive matching degree output submodel is the detail page description matching degree. The comprehensive matching degree output submodel can integrate the outputs of the plurality of previous submodels to finally obtain the final output.
The text-picture matching sub-model, the text-video matching sub-model, the picture-video matching sub-model and the comprehensive matching degree output sub-model are all long-term and short-term neural network models.
And S5, determining the matching degree of the sharing description by using a sharing description matching model based on the information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message.
The shared description matching model is a deep neural network model that includes a deep neural network (Deep Neural Networks, DNN). And the input of the sharing description matching model is information of an item detail page of the shared item in each item sharing message and sharing words corresponding to the shared item in each item sharing message, and the output of the sharing description matching model is the sharing description matching degree.
The matching degree of the sharing description is the matching degree of the information of the commodity detail page of the shared commodity and the sharing text corresponding to the shared commodity. The matching degree of the sharing description can be a numerical value between 0 and 1, the greater the numerical value is, the higher the matching degree of the sharing description is, the more the information of the detail page of the commodity is matched with the corresponding sharing text of the sharing commodity, and the more the information description of the commodity is true. For example, the commodity detail page displays that the commodity is a key mobile phone, and the sharing text corresponding to the shared commodity is a 'large-memory and high-configuration high-end smart mobile phone', so that the sharing description matching degree output by the sharing description matching model is lower, and the sharing description matching degree can be 0.1.
In some embodiments, the method further comprises: and if the matching degree of the sharing description of the commodity sharing message is smaller than the matching threshold of the sharing description, the popup window reminds the shared user to carefully identify the commodity sharing message smaller than the matching threshold of the sharing description. Because if the matching degree of the sharing descriptions is smaller than the matching threshold of the sharing descriptions, the information of the detail pages of the shared commodity is not consistent with the corresponding sharing text of the shared commodity, and the probability of false propaganda of the commodity is high, the user needs to identify carefully.
And S6, constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales volume and the grading information.
In this step, feature information of the commodity node and a relationship between the commodity nodes may be described by constructing the commodity node and a plurality of edges in the commodity node, and the plurality of commodity nodes and the plurality of edges may be input to the graph neural network model to determine an optimal target commodity node. Each commodity node in the commodity nodes comprises a plurality of node characteristics, wherein the node characteristics comprise user similarity, detail page description matching degree, sharing description matching degree, commodity price, commodity type, commodity name, sales volume and grading information of each commodity sharing message. The user similarity, the detail page description matching degree and the sharing description matching degree calculated in the steps S3-S5 can be used as node characteristics of commodity nodes and can be used as input of a follow-up graph neural network model so as to recommend an optimal target commodity node.
The plurality of edges are connecting lines used to represent relationships between the nodes of the commodity. In some embodiments, the characteristics of the edges of the plurality of edges include similarities between commodity nodes. In some embodiments, the method of calculating the similarity between commodity nodes may include histogram matching, perceptual hashing algorithms, and the like.
And S7, processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model to determine a target commodity node, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node.
The graphic neural network model comprises a graphic neural network (Graph Neural Network, GNN) and a full connection layer, the graphic neural network model is an implementation mode of artificial intelligence, the graphic neural network is a neural network directly acting on graphic structure data, and the graphic structure data is a data structure formed by nodes and edges.
The input of the graph neural network model is a plurality of commodity nodes and a plurality of edges between the commodity nodes, and the output of the graph neural network model is the target commodity node. The graph neural network model can be obtained by training marked graph structure data in historical data.
The target commodity node indicates that the commodity node can be recommended to a user for purchase, and the target commodity node is one or more of a plurality of commodity nodes.
And S8, carrying out message reminding on the shared user based on the commodity sharing message corresponding to the target commodity node.
After the target commodity node is determined, the commodity sharing message corresponding to the target commodity node can remind the shared user in the shopping group in a message reminding mode. In some embodiments, the manner of message alert may include a vibration alert, an audible alert, a window alert. For example, the commodity sharing message corresponding to the target commodity node can be directly used as a popup window to prompt the user. For another example, a sound or vibration may be emitted to alert the shared users.
In some embodiments, the method further comprises: and commodity sharing messages corresponding to other commodity nodes which are not selected as target commodity nodes in the commodity nodes can be processed in a mode of only receiving the message and not reminding.
Based on the same inventive concept, fig. 2 is a schematic diagram of an artificial intelligence based message reminding system according to an embodiment of the present invention, where the artificial intelligence based message reminding system includes:
the first obtaining module 21 is configured to obtain a plurality of commodity sharing messages in a shopping group and information of a shared user, where each commodity sharing message in the plurality of commodity sharing messages includes shared commodity information and shared text corresponding to a shared commodity;
the second obtaining module 22 is configured to obtain information of the sharing user corresponding to each commodity sharing message, and information of a commodity detail page of the shared commodity in each commodity sharing message;
the user similarity calculating module 23 is configured to calculate a user similarity based on information of the sharing user corresponding to the commodity sharing message and information of the shared user;
the detail page matching module 24 is configured to determine a detail page description matching degree by using a detail page description matching model based on information of the item detail pages of the shared items in the item sharing message;
the sharing description matching module 25 is configured to determine a sharing description matching degree by using a sharing description matching model based on information of an item detail page of the shared item in the each item sharing message and sharing words corresponding to the shared item in the each item sharing message;
the construction module 26 is configured to construct a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the each commodity sharing message, the user similarity of the each commodity sharing message, the detail page description matching degree of the each commodity sharing message, and the description matching degree of the each commodity sharing message, where the commodity nodes include a plurality of node features, and the plurality of node features include the user similarity of the each commodity sharing message, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales volume, and the scoring information;
a target commodity node determining module 27, configured to determine a target commodity node by processing a plurality of edges between the plurality of commodity nodes based on a graph neural network model, where an input of the graph neural network model is the plurality of edges between the plurality of commodity nodes and the plurality of commodity nodes, and an output of the graph neural network model is the target commodity node;
and the reminding module 28 is used for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node.
Claims (10)
1. An artificial intelligence based message reminding method, which is characterized by comprising the following steps:
acquiring a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared characters corresponding to the shared commodity;
acquiring information of a sharing user corresponding to each commodity sharing message and information of a commodity detail page of the shared commodity in each commodity sharing message;
calculating user similarity based on the information of the sharing users corresponding to each commodity sharing message and the information of the shared users;
determining the detail page description matching degree by using a detail page description matching model based on the information of the commodity detail pages of the shared commodity in each commodity sharing message;
determining the matching degree of the sharing description by using a sharing description matching model based on information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message;
constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales and scoring information;
processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model to determine a target commodity node, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node;
and carrying out message reminding on the shared user based on the commodity sharing message corresponding to the target commodity node.
2. The method for reminding the user of the message based on the artificial intelligence according to claim 1, wherein the calculating the user similarity based on the information of the sharing user corresponding to each commodity sharing message and the information of the shared user comprises: and calculating the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user, and calculating the similarity of the SimHash value of the information of the shared user corresponding to each commodity sharing message and the SimHash value of the information of the shared user according to the Hamming distance.
3. The method for reminding the shared user of the message based on the artificial intelligence according to claim 1, wherein the method for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node comprises vibration reminding, sound reminding and window reminding.
4. The message reminding method based on artificial intelligence according to claim 1, wherein the information of the commodity detail page of the shared commodity in each commodity sharing message comprises video introduction information, picture introduction information and text introduction information.
5. The artificial intelligence based message alert method as claimed in claim 4, wherein the method further comprises: and if the matching degree of the sharing description of the commodity sharing message is smaller than the matching threshold of the sharing description, the popup window reminds the shared user to carefully identify the commodity sharing message smaller than the matching threshold of the sharing description.
6. An artificial intelligence based message alert system comprising:
the first acquisition module is used for acquiring a plurality of commodity sharing messages and information of shared users in a shopping group, wherein each commodity sharing message in the plurality of commodity sharing messages comprises shared commodity information and shared words corresponding to the shared commodity;
the second acquisition module is used for acquiring information of the sharing users corresponding to each commodity sharing message and information of commodity detail pages of the shared commodities in each commodity sharing message;
the user similarity calculation module is used for calculating user similarity based on the information of the sharing users corresponding to each commodity sharing message and the information of the shared users;
the detail page matching module is used for determining the detail page description matching degree by using a detail page description matching model based on the information of the commodity detail pages of the shared commodities in each commodity sharing message;
the sharing description matching module is used for determining the matching degree of the sharing description based on the information of the commodity detail page of the shared commodity in each commodity sharing message and the sharing text corresponding to the shared commodity in each commodity sharing message by using the sharing description matching model;
the construction module is used for constructing a plurality of commodity nodes and a plurality of edges between the commodity nodes based on the commodity sharing messages, the user similarity of the commodity sharing messages, the detail page description matching degree of the commodity sharing messages and the description matching degree of the commodity sharing messages, wherein the commodity nodes comprise a plurality of node characteristics, and the node characteristics comprise the user similarity of the commodity sharing messages, the detail page description matching degree, the sharing description matching degree, the commodity price, the commodity type, the commodity name, the sales and scoring information;
the target commodity node determining module is used for processing the commodity nodes and the edges between the commodity nodes based on a graph neural network model, wherein the input of the graph neural network model is the commodity nodes and the edges between the commodity nodes, and the output of the graph neural network model is the target commodity node;
and the reminding module is used for reminding the shared user of the message based on the commodity sharing message corresponding to the target commodity node.
7. The message reminding system based on artificial intelligence according to claim 6, wherein the user similarity calculation module is further configured to calculate a SimHash value of information of the shared user corresponding to each commodity sharing message and a SimHash value of information of the shared user, and calculate a similarity between the SimHash value of information of the shared user corresponding to each commodity sharing message and the SimHash value of information of the shared user according to a hamming distance.
8. The message alert system based on artificial intelligence according to claim 6, wherein the manner of alerting the shared user based on the commodity sharing message corresponding to the target commodity node includes vibration alert, sound alert, window alert.
9. The artificial intelligence based message alert system according to claim 6, wherein the information of the item detail page of the shared item in each item sharing message includes video introduction information, picture introduction information, text introduction information.
10. The artificial intelligence based message alert system of claim 6, wherein the system is further configured to: and if the matching degree of the sharing description of the commodity sharing message is smaller than the matching threshold of the sharing description, the popup window reminds the shared user to carefully identify the commodity sharing message smaller than the matching threshold of the sharing description.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310884919.0A CN116611897B (en) | 2023-07-19 | 2023-07-19 | Message reminding method and system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310884919.0A CN116611897B (en) | 2023-07-19 | 2023-07-19 | Message reminding method and system based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116611897A CN116611897A (en) | 2023-08-18 |
| CN116611897B true CN116611897B (en) | 2023-10-13 |
Family
ID=87678651
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310884919.0A Active CN116611897B (en) | 2023-07-19 | 2023-07-19 | Message reminding method and system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116611897B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118503807B (en) * | 2024-07-16 | 2024-10-25 | 深圳优特云商贸有限公司 | Multi-dimensional cross-border commodity matching method and system |
| CN119380496B (en) * | 2024-12-27 | 2025-03-21 | 四川省公路规划勘察设计研究院有限公司 | A geological disaster intelligent early warning method and system based on multi-source remote sensing data |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631735A (en) * | 2016-02-05 | 2016-06-01 | 成都梅泰诺移动信息技术有限公司 | Commodity information processing method and device |
| CN105763580A (en) * | 2014-12-15 | 2016-07-13 | 阿里巴巴集团控股有限公司 | Data information sharing method and device |
| CN107169834A (en) * | 2017-05-17 | 2017-09-15 | 丁知平 | A kind of method and apparatus that shopping recommendation is carried out based on big data |
| CN109636545A (en) * | 2018-12-26 | 2019-04-16 | 广州市耀锋电子网络科技有限公司 | A kind of electric business platform commercial product recommending algorithm |
| KR102208418B1 (en) * | 2019-08-12 | 2021-01-27 | 엘지전자 주식회사 | Biometric Apparatus and Method of Sharing Vehicle |
| CN112633978A (en) * | 2020-12-22 | 2021-04-09 | 重庆大学 | Graph neural network model construction method, and method, device and equipment for commodity recommendation |
| CN114037106A (en) * | 2021-09-21 | 2022-02-11 | 图林科技(深圳)有限公司 | Electronic commerce appointment management method and system based on artificial intelligence |
| CN115439139A (en) * | 2022-08-10 | 2022-12-06 | 武汉沁纯服饰有限公司 | User interest analysis method based on E-commerce big data |
| CN115760276A (en) * | 2022-11-03 | 2023-03-07 | 用友网络科技股份有限公司 | Recommendation method, recommendation device and recommendation system based on graph neural network |
| CN115760271A (en) * | 2022-10-28 | 2023-03-07 | 江苏理工学院 | Electromechanical commodity personalized recommendation method and system based on graph neural network |
| CN116204612A (en) * | 2022-10-20 | 2023-06-02 | 超聚变数字技术有限公司 | Text similarity calculation method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11475398B2 (en) * | 2018-05-30 | 2022-10-18 | Ncr Corporation | Product traceability processing |
-
2023
- 2023-07-19 CN CN202310884919.0A patent/CN116611897B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105763580A (en) * | 2014-12-15 | 2016-07-13 | 阿里巴巴集团控股有限公司 | Data information sharing method and device |
| CN105631735A (en) * | 2016-02-05 | 2016-06-01 | 成都梅泰诺移动信息技术有限公司 | Commodity information processing method and device |
| CN107169834A (en) * | 2017-05-17 | 2017-09-15 | 丁知平 | A kind of method and apparatus that shopping recommendation is carried out based on big data |
| CN109636545A (en) * | 2018-12-26 | 2019-04-16 | 广州市耀锋电子网络科技有限公司 | A kind of electric business platform commercial product recommending algorithm |
| KR102208418B1 (en) * | 2019-08-12 | 2021-01-27 | 엘지전자 주식회사 | Biometric Apparatus and Method of Sharing Vehicle |
| CN112633978A (en) * | 2020-12-22 | 2021-04-09 | 重庆大学 | Graph neural network model construction method, and method, device and equipment for commodity recommendation |
| CN114037106A (en) * | 2021-09-21 | 2022-02-11 | 图林科技(深圳)有限公司 | Electronic commerce appointment management method and system based on artificial intelligence |
| CN115439139A (en) * | 2022-08-10 | 2022-12-06 | 武汉沁纯服饰有限公司 | User interest analysis method based on E-commerce big data |
| CN116204612A (en) * | 2022-10-20 | 2023-06-02 | 超聚变数字技术有限公司 | Text similarity calculation method and system |
| CN115760271A (en) * | 2022-10-28 | 2023-03-07 | 江苏理工学院 | Electromechanical commodity personalized recommendation method and system based on graph neural network |
| CN115760276A (en) * | 2022-11-03 | 2023-03-07 | 用友网络科技股份有限公司 | Recommendation method, recommendation device and recommendation system based on graph neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116611897A (en) | 2023-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110909176B (en) | Data recommendation method and device, computer equipment and storage medium | |
| CN108876526B (en) | Product recommendation method, apparatus and computer-readable storage medium | |
| CN110910199B (en) | Method, device, computer equipment and storage medium for ordering project information | |
| CN111784455A (en) | Article recommendation method and recommendation equipment | |
| CN109299994B (en) | Recommendation method, device, equipment and readable storage medium | |
| CN109034973B (en) | Commodity recommendation method, commodity recommendation device, commodity recommendation system and computer-readable storage medium | |
| CN112132660B (en) | Commodity recommendation method, system, equipment and storage medium | |
| US20100106573A1 (en) | Action suggestions based on inferred social relationships | |
| CN116611897B (en) | Message reminding method and system based on artificial intelligence | |
| CN103970850B (en) | Site information recommends method and system | |
| CN106874314B (en) | Information recommendation method and device | |
| CN114862516A (en) | Document recommendation method, storage medium, and program product | |
| CN111225009B (en) | Method and device for generating information | |
| CN108665083A (en) | A kind of method and system for advertisement recommendation for dynamic trajectory model of being drawn a portrait based on user | |
| CN111028029A (en) | An offline product recommendation method, device and electronic device | |
| CN114841760B (en) | Advertisement recommendation management method and system based on audience behavior characteristic analysis | |
| CN113168522B (en) | Server, method and computer-readable storage medium for selecting eyewear equipment | |
| CN113704630B (en) | Information pushing method and device, readable storage medium and electronic equipment | |
| CN117252667A (en) | A product recommendation method and system based on big data | |
| CN111787042B (en) | Method and device for pushing information | |
| CN116308556A (en) | Advertisement pushing method and system based on Internet of things | |
| JP2017167759A (en) | Coordination recommendation device and program | |
| CN121458419B (en) | Product recommendation methods based on big data inference and fine-grained user demand analysis | |
| CN114612142B (en) | Multi-mode information fusion commercial content recommendation method and device and electronic equipment | |
| US20230153883A1 (en) | Methods and apparatus for determining attribute affinities for users |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230915 Address after: 3181, 3rd Floor, Building 2, China Agricultural University International Entrepreneurship Park, No. 10 Tianxiu Road, Haidian District, Beijing, 100091 Applicant after: Beijing Kuaiyitong Technology Co.,Ltd. Address before: 644000 floor 6, building 1, science and innovation building, industrial headquarters base, No. 9, shaping Road, Guoxing Avenue, Lingang Economic Development Zone, Yibin City, Sichuan Province Applicant before: Yibin Xukong Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |