CN115187343B - Attention graph convolution neural network-based multi-behavior recommendation method - Google Patents
Attention graph convolution neural network-based multi-behavior recommendation method Download PDFInfo
- Publication number
- CN115187343B CN115187343B CN202210850345.0A CN202210850345A CN115187343B CN 115187343 B CN115187343 B CN 115187343B CN 202210850345 A CN202210850345 A CN 202210850345A CN 115187343 B CN115187343 B CN 115187343B
- Authority
- CN
- China
- Prior art keywords
- user
- behavior
- kth
- matrix
- ith
- 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; CALCULATING OR 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—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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]
-
- 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
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A multi-behavior recommendation method based on an attention graph convolution neural network utilizes a multi-channel graph convolution neural network module to design a characteristic expression for each behavior so as to solve the problem of multiple embedded expressions under multiple behaviors; capturing the importance degrees of different behaviors by adopting a behavior special attention mechanism module; finally modeling the user preference through the multi-task learning framework, improving the recommendation effect of the model, and recommending through modeling the user preference.
Description
Technical Field
The invention relates to the technical field of recommendation systems and deep learning, in particular to a multi-behavior recommendation method based on an attention map convolution neural network.
Background
In the age of information explosion, the recommendation system plays a great role in solving the problem of information overload. The recommendation system can predict new information of interest to the user from historical information of the user in the past, and is widely applied to platforms such as music, movies, electronic commerce, online comment websites, location-based recommendation services and the like. Among them, collaborative Filtering (CF) technology is the most widely used algorithm in the recommended field. However, most of the conventional models and the neural network models are used for researching single-type behaviors, and interests of users in real life are not only reflected on the single behaviors, so that research on multi-behavior recommendation is necessary. The current model study for multi-behavioral recommendations is based mainly on these four frameworks: BPR, RNN, GCN, MTL. The BPR framework is used for sampling a plurality of lines of data, and the auxiliary behavior is used for ordering and serving the object pairs of the user; the RNN network is utilized to capture sequence information of various behaviors, and a special attention mechanism of the behaviors is utilized to solve the importance degree among the behaviors. The GCN network is utilized to capture multiple behavior information of the user and item map and item attribute information of the item and item map. The MTL framework is utilized to enable a plurality of tasks to share the user to be embedded or to conduct model training according to the sequence relation among behaviors. The above work, while achieving great success, ignores the association between different embedded expressions and failure to mine deeper into the behavior under a variety of behaviors, resulting in a less than optimal model.
Disclosure of Invention
In order to overcome the defects of the technology, the invention provides a method for modeling user preference and improving the recommending effect of a model by constructing an embedded expression of various behaviors, a multi-channel graph convolution module and adopting a behavior special attention mechanism module.
The technical scheme adopted for overcoming the technical problems is as follows:
a multi-behavior recommendation method based on an attention graph convolution neural network comprises the following steps:
a) A leave-one method is adopted in the e-commerce data set to obtain a training set and a testing set;
b) The training set comprises N users, I articles and K types of user behaviors, and an undirected graph of the user behaviors is constructed by utilizing the training set;
c) Establishing and training a multitask learning network model based on a behavior-specific attention multichannel graph convolutional nerve;
d) And calculating the preference degree of the user for performing certain action on the article, and realizing article recommendation.
Further, each user in the test set in step a) has an item that has not interacted with it, and parameters of the multitask learning network model based on the behavior-specific attention multi-channel graph convolutional nerves in step c) are debugged by the test set.
Further, in step b), the k=3, 3 types of user behaviors are the user's purchase of an item, the user's addition of an item to a shopping cart, and the user's click on an item link, respectively.
Further, step b) comprises the steps of:
b-1) is represented by formula g= (U) (k) ,I (k) ) Constructing an undirected graph G, k= {1,2,..once, K } of the kth user behavior, wherein U (k) In order to contain the set of user nodes,wherein->Is the vector matrix of the nth user under the kth user behavior, n= {1,2, …, N }, +.>Wherein->Vector matrix for the nth item under kth user behavior, m= {1,2, …, I }, +.>Where R is the real space and d is the dimension. Further, step c) comprises the steps of:
c-1) by the formulaComputing the embedded expression matrix of the u-th user at the first layer of the undirected graph G +.>In->The embedded expression matrix of item i for layer l-1 of undirected graph G, l= {1,2,3}, N u For the number of items interacted with by the u-th user, N i An amount of interaction by the ith item by the ith user;
c-2) passing through a maleA kind of electronic device with high-pressure air-conditioning systemCalculating the embedded expression matrix of the ith item in the first layer of the undirected graph G +.>In->An embedded expression matrix of a user u when the embedded expression matrix is the l-1 layer of the undirected graph G;
c-3) passing through the formulaCalculating to obtain an embedded expression matrix e of a final u-th user after l layers are aggregated under each behavior u ;
c-4) passing through the formulaCalculating to obtain an embedded expression matrix e of the final ith article after l layers are aggregated under each behavior i ;
c-5) passing through the formulaCalculating to obtain an embedded expression matrix of the u user under the kth action>In->Initializing a matrix for the weight of the kth user affected by other actions in the kth action, b is a bias initialization factor, +.>Initializing a matrix for embedded expressions of the u-th user before being affected by the k-th behavior;
c-6) passing through the formulaCalculating to obtain the embedded expression matrix of the ith article under the kth action>In->Initializing a matrix for the weight of the ith item affected by other actions in the kth action, b is a bias initialization factor, +.>Initializing a matrix for embedding expression of an ith article before being influenced by a kth behavior, and completing establishment of a multitask learning network model based on behavior-specific attention multichannel graph convolutional nerves;
c-7) passing through the formulaCalculating loss L of kth behavior k Wherein sigma is a sigmoid function, lambda is an L2 norm normalization factor, and ++>Possibility of performing kth behavior on the ith item for the ith user,/-> Possibility of performing kth action on the jth item for the jth user,/->j={1,2,…,I},
c-8) passing through a maleA kind of electronic device with high-pressure air-conditioning systemCalculating to obtain total loss L MTL Wherein a is k Loss weight value for kth behavior;
c-9) utilizing the total loss L MTL Training a multitask learning network model based on a behavior special attention multichannel graph convolutional nerve by adopting a gradient descent method, and updating an embedded expression matrix of a ith user under a kth behavior in the step c-5) through back propagationAnd a weight initialization matrix for the kth user affected by other actions in the kth action +.>Updating the embedded expression matrix of the ith item in step c-6) under the kth action +.>And a weight initialization matrix for the ith item under the kth action affected by other actions +.>
Further, in step d) the formula is passedAnd calculating to obtain the preference degree y of the user on the article, wherein T is the transposition.
The beneficial effects of the invention are as follows: designing a characteristic expression for each behavior by utilizing a multichannel graph convolution neural network module to solve the problem of multiple embedded expressions under multiple behaviors; capturing the importance degrees of different behaviors by adopting a behavior special attention mechanism module; finally modeling the user preference through the multi-task learning framework, improving the recommendation effect of the model, and recommending through modeling the user preference.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a block diagram of a behavior-tailored attention mechanism of the present invention.
Detailed Description
The invention is further described with reference to fig. 1 and 2.
A multi-behavior recommendation method based on an attention graph convolution neural network comprises the following steps:
a) A leave-one method is adopted in the e-commerce data set to obtain a training set and a testing set;
b) The training set comprises N users, I articles and K types of user behaviors, and an undirected graph of the user behaviors is constructed by utilizing the training set;
c) Establishing and training a multitask learning network model based on a behavior-specific attention multichannel graph convolutional nerve;
d) And calculating the preference degree of the user for performing certain action on the article, and realizing article recommendation.
Designing a characteristic expression for each behavior by utilizing a multichannel graph convolution neural network module to solve the problem of multiple embedded expressions under multiple behaviors; capturing the importance degrees of different behaviors by adopting a behavior special attention mechanism module; finally modeling the user preference through the multi-task learning framework, improving the recommendation effect of the model, and recommending through modeling the user preference.
Example 1:
each user in the test set in step a) has an item that has not interacted with it, and the parameters of the multitask learning network model based on the behavior-specific attention-multichannel graph convolutional nerves in step c) are debugged by the test set.
Example 2:
the k=3, 3 types of user actions in step b) are the user's purchase of items, the user's addition of items to the shopping cart, and the user's click on item links, respectively. For example, a real e-commerce data set is used, which is from the largest mother and infant e-commerce platform shellfish in China, and comprises 3 types of user behaviors, including user purchasing behavior, user adding the object to a shopping cart behavior and user clicking object linking behavior.
Example 3:
step b) comprises the steps of:
b-1) is represented by formula g= (U) (k) ,I (k) ) Constructing a kth undirected graph G of user behaviors, wherein k= {1,2, …, K }, k=1 is a user purchasing article behavior, k=2 is a user adding article to shopping cart behavior, and k=3 is a user clicking article linking behavior, wherein U is as follows (k) In order to contain the set of user nodes,wherein->Is the vector matrix of the nth user under the kth user behavior, n= {1,2, …, N }, +.>Wherein->Vector matrix for the nth item under kth user behavior, m= {1,2, …, I }, +.>Where R is the real space and d is the dimension.
Example 4:
step c) comprises the steps of:
c-1) by the formulaComputing the embedded expression matrix of the u-th user at the first layer of the undirected graph G +.>In->As undirected graph GEmbedding expression matrix of object i in l-1 layer, l= {1,2,3}, N u For the number of items interacted with by the u-th user, N i Is the number of items that the ith item interacted with by the ith user.
c-2) passing through the formulaCalculating the embedded expression matrix of the ith item in the first layer of the undirected graph G +.>In->The expression matrix is embedded by user u at layer l-1 of undirected graph G.
c-3) passing through the formulaCalculating to obtain an embedded expression matrix e of a final u-th user after l layers are aggregated under each behavior u 。
c-4) passing through the formulaCalculating to obtain an embedded expression matrix e of the final ith article after l layers are aggregated under each behavior i 。
c-5) passing through the formulaCalculating to obtain an embedded expression matrix of the u user under the kth action>In->Initializing a matrix (trainable) for the weight of the kth user affected by other actions at the kth action, b being a bias initialization factor (a trainable real number),>the matrix is initialized for the embedded representation of the u-th user before being affected by the kth behavior. c-6) by the formula->Calculating to obtain the embedded expression matrix of the ith article under the kth action>In->Initializing a matrix (trainable) for weights of the ith item affected by other actions at the kth action, b is a bias initialization factor (a trainable real number),initializing a matrix for embedding expression of the ith article before being affected by the kth behavior, and completing establishment of a multitask learning network model based on behavior-specific attention multichannel graph convolutional nerves.
c-7) passing through the formulaCalculating loss L of kth behavior k Wherein sigma is a sigmoid function, lambda is an L2 norm normalization factor, and ++>Possibility of performing kth behavior on the ith item for the ith user,/->i={1,2,…,I},/>Possibility of performing kth action on the jth item for the jth user,/->j={1,2,...,I},
c-8) passing through the formulaCalculating to obtain total loss L MTL Wherein a is k Is the loss weight value of the kth behavior.
c-9) utilizing the total loss L MTL Training a multitask learning network model based on a behavior special attention multichannel graph convolutional nerve by adopting a gradient descent method, and updating an embedded expression matrix of a ith user under a kth behavior in the step c-5) through back propagationAnd a weight initialization matrix for the kth user affected by other actions in the kth action +.>Updating the embedded expression matrix of the ith item in step c-6) under the kth action +.>And a weight initialization matrix for the ith item under the kth action affected by other actions +.>
Example 5:
in step d) by the formulaAnd calculating to obtain the preference degree y of the user on the article, wherein T is the transposition. Finally, it should be noted that: the above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, although the present invention has been described in detail with reference to the foregoing embodimentsIt will be apparent to those skilled in the art that modifications may be made to the embodiments described above, or equivalents may be substituted for elements thereof. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention. />
Claims (5)
1. The multi-behavior recommendation method based on the attention map convolution neural network is characterized by comprising the following steps of:
a) A leave-one method is adopted in the e-commerce data set to obtain a training set and a testing set;
b) The training set comprises N users, I articles and K types of user behaviors, and an undirected graph of the user behaviors is constructed by utilizing the training set;
c) Establishing and training a multitask learning network model based on a behavior-specific attention multichannel graph convolutional nerve;
d) Calculating the preference degree of a user for performing certain action on the article, and realizing article recommendation;
step c) comprises the steps of:
c-1) by the formulaComputing the embedded expression matrix of the u-th user at the first layer of the undirected graph G +.>In->The embedded expression matrix of item i for layer l-1 of undirected graph G, l= {1,2,3}, N u For the number of items interacted with by the u-th user, N i An amount of interaction by the ith item by the ith user;
c-2) passing through the formulaCalculating the embedded expression of the ith item at the first layer of the undirected graph GMatrix->In->An embedded expression matrix of a user u when the embedded expression matrix is the l-1 layer of the undirected graph G;
c-3) passing through the formulaCalculating to obtain an embedded expression matrix e of a final u-th user after l layers are aggregated under each behavior u ;
c-4) passing through the formulaCalculating to obtain an embedded expression matrix e of the final ith article after l layers are aggregated under each behavior i ;
c-5) passing through the formulaCalculating to obtain an embedded expression matrix of the u user under the kth action>In->Initializing a matrix for the weight of the kth user affected by other actions in the kth action, b is a bias initialization factor, +.>Initializing a matrix for embedded expressions of the u-th user before being affected by the k-th behavior;
c-6) passing through the formulaCalculating to obtain the embedded expression matrix of the ith article under the kth action>W in the formula i (k) Initializing a matrix for the weight of the ith item affected by other actions in the kth action, b is a bias initialization factor, +.>Initializing a matrix for embedding expression of an ith article before being influenced by a kth behavior, and completing establishment of a multitask learning network model based on behavior-specific attention multichannel graph convolutional nerves;
c-7) passing through the formulaCalculating loss L of kth behavior k Wherein sigma is a sigmoid function, lambda is an L2 norm normalization factor, and ++>Possibility of performing kth behavior on the ith item for the ith user,/-> Possibility of performing kth action on the jth item for the jth user,/->
c-8) passing through the formulaCalculating to obtain total loss L MTL Wherein a is k Loss weight value for kth behavior;
c-9) utilizing the total loss L MTL Training based on behavior-tailored attention multi-channel using gradient descent methodThe multitask learning network model of the graph convolutional nerve updates the embedded expression matrix of the ith user under the kth action in the step c-5) through back propagationAnd a weight initialization matrix for the kth user affected by other actions in the kth action +.>Updating the embedded expression matrix of the ith item in step c-6) under the kth action +.>And a weight initialization matrix W for the ith item under the kth behavior affected by other behaviors i (k) 。
2. The attention-map-based convolutional neural network-based multi-behavior recommendation method of claim 1, wherein: each user in the test set in step a) has an item not interacted with it, and the parameters of the multitask learning network model based on the behavior-specific attention multi-channel graph convolutional nerves in step c) are debugged through the test set.
3. The attention-map-based convolutional neural network-based multi-behavior recommendation method of claim 1, wherein: the k=3, 3 types of user actions in step b) are the user's purchase of items, the user's addition of items to the shopping cart, and the user's click on item links, respectively.
4. The attention map convolution neural network based multi-behavior recommendation method according to claim 1, wherein the step b) includes the steps of:
b-1) is represented by formula g= (U) (k) ,I (k) ) Constructing an undirected graph G, k= {1,2, …, K } of kth user behavior, wherein U is as follows (k) In order to contain the set of user nodes,wherein->Is the vector matrix of the nth user under the kth user behavior, n= {1,2, …, N }, +.>Wherein->Vector matrix for the nth item under kth user behavior, m= {1,2, …, I }, +.>Where R is the real space and d is the dimension.
5. The attention-map-based convolutional neural network-based multi-behavior recommendation method of claim 1, wherein: in step d) by the formulaAnd calculating to obtain the preference degree y of the user on the article, wherein T is the transposition.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210850345.0A CN115187343B (en) | 2022-07-20 | 2022-07-20 | Attention graph convolution neural network-based multi-behavior recommendation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210850345.0A CN115187343B (en) | 2022-07-20 | 2022-07-20 | Attention graph convolution neural network-based multi-behavior recommendation method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115187343A CN115187343A (en) | 2022-10-14 |
CN115187343B true CN115187343B (en) | 2023-08-08 |
Family
ID=83519039
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210850345.0A Active CN115187343B (en) | 2022-07-20 | 2022-07-20 | Attention graph convolution neural network-based multi-behavior recommendation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115187343B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116645174B (en) * | 2023-07-27 | 2023-10-17 | 山东省人工智能研究院 | Personalized recommendation method based on decoupling multi-behavior characterization learning |
CN117171448A (en) * | 2023-08-11 | 2023-12-05 | 哈尔滨工业大学 | Multi-behavior socialization recommendation method and system based on graph neural network |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110580314A (en) * | 2019-07-16 | 2019-12-17 | 华南师范大学 | Course recommendation method and system based on graph convolution neural network and dynamic weight |
CN110674407A (en) * | 2019-09-30 | 2020-01-10 | 杭州电子科技大学 | Hybrid recommendation method based on graph convolution neural network |
CN111127142A (en) * | 2019-12-16 | 2020-05-08 | 东北大学秦皇岛分校 | Article recommendation method based on generalized neural attention |
CN111523047A (en) * | 2020-04-13 | 2020-08-11 | 中南大学 | Multi-relation collaborative filtering algorithm based on graph neural network |
CN111737569A (en) * | 2020-06-04 | 2020-10-02 | 山东省人工智能研究院 | Personalized recommendation method based on attribute perception intention-minded convolutional neural network |
CN112232925A (en) * | 2020-11-02 | 2021-01-15 | 哈尔滨工程大学 | Method for carrying out personalized recommendation on commodities by fusing knowledge maps |
CN114117220A (en) * | 2021-11-26 | 2022-03-01 | 东北大学 | Deep reinforcement learning interactive recommendation system and method based on knowledge enhancement |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111581520B (en) * | 2020-05-25 | 2022-04-19 | 中国人民解放军国防科技大学 | Item recommendation method and system based on item importance in session |
-
2022
- 2022-07-20 CN CN202210850345.0A patent/CN115187343B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110580314A (en) * | 2019-07-16 | 2019-12-17 | 华南师范大学 | Course recommendation method and system based on graph convolution neural network and dynamic weight |
CN110674407A (en) * | 2019-09-30 | 2020-01-10 | 杭州电子科技大学 | Hybrid recommendation method based on graph convolution neural network |
CN111127142A (en) * | 2019-12-16 | 2020-05-08 | 东北大学秦皇岛分校 | Article recommendation method based on generalized neural attention |
CN111523047A (en) * | 2020-04-13 | 2020-08-11 | 中南大学 | Multi-relation collaborative filtering algorithm based on graph neural network |
CN111737569A (en) * | 2020-06-04 | 2020-10-02 | 山东省人工智能研究院 | Personalized recommendation method based on attribute perception intention-minded convolutional neural network |
CN112232925A (en) * | 2020-11-02 | 2021-01-15 | 哈尔滨工程大学 | Method for carrying out personalized recommendation on commodities by fusing knowledge maps |
CN114117220A (en) * | 2021-11-26 | 2022-03-01 | 东北大学 | Deep reinforcement learning interactive recommendation system and method based on knowledge enhancement |
Non-Patent Citations (1)
Title |
---|
基于注意力机制增强图卷积神经网络的个性化新闻推荐;杨宝生;;兰州文理学院学报(自然科学版)(第05期);第52-56页 * |
Also Published As
Publication number | Publication date |
---|---|
CN115187343A (en) | 2022-10-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115187343B (en) | Attention graph convolution neural network-based multi-behavior recommendation method | |
CN111428147B (en) | Social recommendation method of heterogeneous graph volume network combining social and interest information | |
CN108874914A (en) | A kind of information recommendation method based on the long-pending and neural collaborative filtering of picture scroll | |
CN106920147A (en) | A kind of commodity intelligent recommendation method that word-based vector data drives | |
CN103514255B (en) | A kind of collaborative filtering recommending method based on project stratigraphic classification | |
CN111881342A (en) | Recommendation method based on graph twin network | |
CN106776928B (en) | Position recommendation method based on memory computing framework and fusing social contact and space-time data | |
CN111259263B (en) | Article recommendation method and device, computer equipment and storage medium | |
CN108664658A (en) | A kind of collaborative filtering video recommendation method considering user preference dynamic change | |
Mousavi et al. | Adaptive sequentially space‐filling metamodeling applied in optimal water quantity allocation at basin scale | |
CN112416986B (en) | User portrait realizing method and system based on hierarchical personalized federal learning | |
CN112529415B (en) | Article scoring method based on combined multiple receptive field graph neural network | |
CN103353880A (en) | Data mining method adopting dissimilarity degree clustering and association | |
CN111737569B (en) | Personalized recommendation method based on attribute perception intention-minded convolutional neural network | |
CN113868537B (en) | Recommendation method based on multi-behavior session graph fusion | |
CN110390056A (en) | Big data processing method, device, equipment and readable storage medium storing program for executing | |
CN112765482A (en) | Product delivery method, device, equipment and computer readable medium | |
CN114385150A (en) | Low code development method based on intelligent engine | |
CN109087146A (en) | The prediction technique and system of movie theatre box-office income | |
CN110570226B (en) | Scoring prediction method combining topic model and heterogeneous information network | |
Hong | LSTM-based Sales Forecasting Model. | |
CN115309997B (en) | Commodity recommendation method and device based on multi-view self-coding features | |
CN112016000A (en) | Movie group recommendation method and system based on convolution collaborative filtering | |
CN111652695B (en) | Collaborative filtering recommendation method based on parallel self-encoding machine | |
CN115660147A (en) | Information propagation prediction method and system based on influence modeling between propagation paths and in propagation paths |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |