CN110544131A - Data-driven E-commerce user purchasing behavior prediction method - Google Patents
Data-driven E-commerce user purchasing behavior prediction method Download PDFInfo
- Publication number
- CN110544131A CN110544131A CN201910843151.6A CN201910843151A CN110544131A CN 110544131 A CN110544131 A CN 110544131A CN 201910843151 A CN201910843151 A CN 201910843151A CN 110544131 A CN110544131 A CN 110544131A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- feature
- model
- commodity
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 25
- 230000004927 fusion Effects 0.000 claims abstract description 24
- 238000012549 training Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000012216 screening Methods 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- 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]
Abstract
The invention provides a data-driven electric commercial user purchasing behavior prediction method, which comprises the following steps: s1) preprocessing data; s2) feature engineering; s3) data division; s4) feature screening; s5) model training and model fusion; s6) model prediction. After the feature importance degrees are sequenced based on the XGB OST algorithm, the problem of collinearity among features is considered, and features are grouped by taking the Pearson correlation coefficient among the features as a criterion, so that any two feature correlation coefficients in each feature group are less than 0.5; and vertically segmenting the data set by using the feature set after clustering, independently predicting a model, and then performing model fusion. The method completely retains all characteristic information while solving the characteristic co-linearity problem, and reduces the model prediction variance.
Description
Technical Field
The invention relates to the technical field of recommendation systems, in particular to a data-driven electric commercial user purchasing behavior prediction method.
background
when the existing user behavior prediction method faces the characteristics with strong correlation, the model result is deteriorated due to the conflict among the characteristics, and the loss of the characteristic information is caused if the characteristics with strong correlation are directly deleted. Meanwhile, the existing model fusion method only aims at models with different parameters and different types, and the training data set has no difference, so that the difference between the models is insufficient, and the fusion effect cannot be obviously improved.
disclosure of Invention
the invention provides a data-driven electric commerce user purchasing behavior prediction method aiming at key problems in accurate marketing in social electric commerce.
in order to solve the technical problems, the technical scheme of the invention is as follows:
a data-driven electric business user purchasing behavior prediction method comprises the following steps:
s1) data preprocessing: preprocessing the original data set to remove noise data;
S2) feature engineering: constructing a feature set based on data such as service needs, user attributes and commodity information;
s3) data partitioning: the method comprises the steps of horizontally dividing original data into a training set, a verification set and a test set;
s4) feature screening: constructing a prediction target based on the purchasing behavior of the user, and further performing feature filtering according to the relationship between the features and the target;
S5) model training and model fusion: analyzing the performance of the model on the verification set, adjusting the model hyperparameter, and performing multi-model fusion;
s6) model prediction: and predicting the purchasing behavior of the E-commerce user by using the fusion model.
In the above technical solution, the step S1) of data preprocessing specifically includes the following steps:
S11) pulling the user exposure data record, the user click data record and the user purchase data record in the last week, and eliminating data containing abnormal values;
S12) eliminating data which are not exposed records and click records and only purchase records;
s13) to eliminate inactive user data that did not generate a click or purchase record within the last week.
in the above technical solution, step S2) feature engineering specifically includes the following steps:
s21), constructing a user basic attribute feature group, including but not limited to user age, user member level, user member type and user registration area;
s22) constructing basic attribute feature groups of the commodities, including but not limited to commodity type id, commodity hang tag price, discount price and commodity sale price;
s23), constructing a basic statistical characteristic group, including but not limited to the user purchase commodity price preference (such as: maximum, minimum, mean, standard deviation, 25% quantile, median, 75% quantile for all purchased commodity prices for the user's week, clicks and ordering preferences for different categories for the user's week (e.g.: the ratio of the click volume of different categories to the total click volume, the ratio of the purchase quantity of different categories to the total purchase volume, the purchase click ratio of different categories to the exposure click ratio of different categories, the commodity price binning, the number of clicked commodities in a week, the number of exposed commodities, the number of purchased commodities, the number of commodity purchases in the last day of the user, the number of exposed commodities, the number of purchased commodities, the click and order preference of different categories in the last day of the user (such as: the ratio of the different-category click rate to the total click rate, the ratio of the purchase quantity of the different-category commodities to the total purchase quantity, the purchase click rate of the different-category commodities, and the exposure click rate of the different-category commodities).
In the above technical solution, the step S4) of feature screening specifically includes the following steps:
S41) training the features by using an XGB OST algorithm, giving feature importance degree sequencing based on splitting gain, and removing feature sets with smaller feature importance;
S42), in the preserved features, calculating the Pearson correlation coefficient among the features, and grouping the feature sets, wherein the features with larger correlation coefficients are grouped into different groups, so that the correlation coefficient of any two features in each group is less than 0.5.
in the above technical solution, step S5) of model training and model fusion specifically includes the following steps:
S51) vertically segmenting the data set according to the feature grouping result, and dividing the data set into a plurality of data sets;
s52) carrying out XGB OST user purchasing behavior prediction based on different data sets, and respectively carrying out model hyperparameter adjustment on the different data sets;
s53) data set and parameter difference lead to difference between models, and the results of different models are used as input for weighted fusion;
s54) taking AUC as an objective function, taking each model weight as a decision variable, solving the optimal weight vector corresponding to the maximum AUC, and fusing all models.
the technical scheme of the invention has the following beneficial effects:
1. according to the method, the characteristics are clustered based on the linear correlation coefficient among the characteristics, the data set is vertically divided according to the characteristic clustering result, all characteristic information is kept as far as possible, and the model difference is increased, so that a better model fusion effect is obtained.
2. after the importance degrees of the features are sequenced based on the XGB OST algorithm, the problem of collinearity among the features is considered, and the features are grouped by taking the Pearson correlation coefficient among the features as a criterion, so that any two feature correlation coefficients in each feature group are smaller than 0.5.
3. and vertically segmenting the data set by using the feature set after clustering, independently predicting a model, and then performing model fusion. The method completely retains all characteristic information while solving the characteristic co-linearity problem, and reduces the model prediction variance.
4. the method creatively provides a feature clustering method by preprocessing data such as user attribute information, commodity attribute information, user-commodity interaction information and the like and extracting features by combining service requirements, and completely reserves all feature information while solving feature conflicts. Meanwhile, the difference between the models is increased through the splitting of the data set and the adjustment of the parameters, the overfitting risk in the later-stage model fusion is reduced, and the model prediction variance is reduced.
drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a schematic diagram of the overall steps of the technical solution of the present invention.
Detailed Description
As shown in fig. 1, the present invention provides a data-driven method for predicting the purchasing behavior of an e-commerce consumer, comprising the following steps:
s1) data preprocessing: preprocessing the original data set to remove noise data;
s2) feature engineering: constructing a feature set based on data such as service needs, user attributes and commodity information;
S3) data partitioning: the method comprises the steps of horizontally dividing original data into a training set, a verification set and a test set;
s4) feature screening: constructing a prediction target based on the purchasing behavior of the user, and further performing feature filtering according to the relationship between the features and the target;
S5) model training and model fusion: analyzing the performance of the model on the verification set, adjusting the model hyperparameter, and performing multi-model fusion;
s6) model prediction: and predicting the purchasing behavior of the E-commerce user by using the fusion model.
in the above technical solution, the step S1) of data preprocessing specifically includes the following steps:
S11) pulling the user exposure data record, the user click data record and the user purchase data record in the last week, and eliminating data containing abnormal values;
s12) eliminating data which are not exposed records and click records and only purchase records;
S13) to eliminate inactive user data that did not generate a click or purchase record within the last week.
in the above technical solution, step S2) feature engineering specifically includes the following steps:
S21), constructing a user basic attribute feature group, including but not limited to user age, user member level, user member type and user registration area;
s22) constructing basic attribute feature groups of the commodities, including but not limited to commodity type id, commodity hang tag price, discount price and commodity sale price;
S23) constructing a basic statistical feature group:
the above feature groups include, but are not limited to, user purchase price preferences such as: maximum value, minimum value, average value, standard deviation, 25% quantile, median and 75% quantile of all the prices of the purchased commodities in one week;
User click and order preference for different categories within a week, such as: the occupation ratio of the click rate of different types in the total click rate, the occupation ratio of the purchase quantity of different types of commodities in the total purchase quantity, the purchase click ratio of different types of commodities and the exposure click ratio of different types of commodities;
the commodity price is classified into boxes, the number of commodities clicked by the user in one week, the number of exposed commodities, the number of purchased commodities, the number of clicks and ordering preferences of the user on different categories in the last day are as follows: the proportion of different types of click rate in the total click rate, the proportion of different types of commodity purchase quantity in the total purchase quantity, the purchase click rate in different types of commodities and the exposure click rate in different types of commodities.
In the above technical solution, the step S4) of feature screening specifically includes the following steps:
s41) training the features by using an XGB OST algorithm, giving feature importance degree sequencing based on splitting gain, and removing feature sets with smaller feature importance;
S42), in the preserved features, calculating the Pearson correlation coefficient among the features, and grouping the feature sets, wherein the features with larger correlation coefficients are grouped into different groups, so that the correlation coefficient of any two features in each group is less than 0.5.
in the above technical solution, step S5) of model training and model fusion specifically includes the following steps:
S51) vertically segmenting the data set according to the feature grouping result, and dividing the data set into a plurality of data sets;
s52) carrying out XGB OST user purchasing behavior prediction based on different data sets, and respectively carrying out model hyperparameter adjustment on the different data sets;
s53) data set and parameter difference lead to difference between models, and the results of different models are used as input for weighted fusion;
S54) taking AUC as an objective function, taking each model weight as a decision variable, solving the optimal weight vector corresponding to the maximum AUC, and fusing all models.
According to the method, the characteristics are clustered based on the linear correlation coefficient among the characteristics, the data set is vertically divided according to the characteristic clustering result, all characteristic information is kept as far as possible, and the model difference is increased, so that a better model fusion effect is obtained. After the importance degrees of the features are sequenced based on the XGB OST algorithm, the problem of collinearity among the features is considered, and the features are grouped by taking the Pearson correlation coefficient among the features as a criterion, so that the correlation coefficient of any two features in each feature group is less than 0.5. The data set is vertically segmented by using the feature set after clustering, model prediction is independently performed, and then model fusion is performed. The method completely retains all characteristic information while solving the characteristic co-linearity problem, and reduces the model prediction variance. The method creatively provides a feature clustering method by preprocessing data such as user attribute information, commodity attribute information, user-commodity interaction information and the like and extracting features by combining service requirements, and completely reserves all feature information while solving feature conflicts. Meanwhile, the difference between the models is increased through the splitting of the data set and the adjustment of the parameters, the overfitting risk in the later-stage model fusion is reduced, and the model prediction variance is reduced.
the foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (5)
1. a data-driven electric business user purchasing behavior prediction method is characterized by comprising the following steps:
S1) data preprocessing: preprocessing the original data set to remove noise data;
s2) feature engineering: constructing a feature set based on data such as service needs, user attributes and commodity information;
S3) data partitioning: the method comprises the steps of horizontally dividing original data into a training set, a verification set and a test set;
s4) feature screening: constructing a prediction target based on the purchasing behavior of the user, and further performing feature filtering according to the relationship between the features and the target;
S5) model training and model fusion: analyzing the performance of the model on the verification set, adjusting the model hyperparameter, and performing multi-model fusion;
s6) model prediction: and predicting the purchasing behavior of the E-commerce user by using the fusion model.
2. the data-driven electric business user purchasing behavior prediction method as claimed in claim 1, wherein the step S1) of data preprocessing specifically includes the following steps:
s11) pulling the user exposure data record, the user click data record and the user purchase data record in the last week, and eliminating data containing abnormal values;
S12) eliminating data which are not exposed records and click records and only purchase records;
s13) to eliminate inactive user data that did not generate a click or purchase record within the last week.
3. the method for predicting the purchasing behavior of the data-driven electric power business user according to claim 1, wherein the step S2) of feature engineering specifically comprises the following steps:
s21), constructing a user basic attribute feature group, including but not limited to user age, user member level, user member type and user registration area;
S22) constructing basic attribute feature groups of the commodities, including but not limited to commodity type id, commodity hang tag price, discount price and commodity sale price;
s23), constructing a basic statistical characteristic group, including commodity purchase price preference of a user, click and order placing preference of different categories within one week of the user, commodity price binning, commodity number clicked within one week of the user, exposure commodity number, commodity purchase number within the last day of the user, exposure commodity number, commodity purchase number, click and order placing preference of different categories within the last day of the user.
4. the data-driven electric business user purchasing behavior prediction method as claimed in claim 1, wherein the step S4) feature screening specifically includes the following steps:
s41) training the features by using an XGB OST algorithm, giving feature importance degree sequencing based on splitting gain, and removing feature sets with smaller feature importance;
S42), in the preserved features, calculating the Pearson correlation coefficient among the features, and grouping the feature sets, wherein the features with larger correlation coefficients are grouped into different groups, so that the correlation coefficient of any two features in each group is less than 0.5.
5. The data-driven electric business user purchasing behavior prediction method according to claim 1, wherein the step S5) of model training and model fusion specifically comprises the following steps:
S51) vertically segmenting the data set according to the feature grouping result, and dividing the data set into a plurality of data sets;
s52) carrying out XGB OST user purchasing behavior prediction based on different data sets, and respectively carrying out model hyperparameter adjustment on the different data sets;
S53) data set and parameter difference lead to difference between models, and the results of different models are used as input for weighted fusion;
s54) taking AUC as an objective function, taking each model weight as a decision variable, solving the optimal weight vector corresponding to the maximum AUC, and fusing all models.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910843151.6A CN110544131A (en) | 2019-09-06 | 2019-09-06 | Data-driven E-commerce user purchasing behavior prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910843151.6A CN110544131A (en) | 2019-09-06 | 2019-09-06 | Data-driven E-commerce user purchasing behavior prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110544131A true CN110544131A (en) | 2019-12-06 |
Family
ID=68713036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910843151.6A Pending CN110544131A (en) | 2019-09-06 | 2019-09-06 | Data-driven E-commerce user purchasing behavior prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110544131A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112270545A (en) * | 2020-10-27 | 2021-01-26 | 上海淇馥信息技术有限公司 | Financial risk prediction method and device based on migration sample screening and electronic equipment |
| CN113283960A (en) * | 2021-05-18 | 2021-08-20 | 武汉畅来电子商务有限公司 | Vertical e-commerce platform commodity intelligent recommendation method based on big data analysis and cloud computing and cloud service platform |
| CN113592160A (en) * | 2021-07-13 | 2021-11-02 | 北京淇瑀信息科技有限公司 | Fusion processing-based loss of contact risk prediction method and system for user equipment and computer equipment |
| CN114511022A (en) * | 2022-01-24 | 2022-05-17 | 百度在线网络技术（北京）有限公司 | Feature screening, behavior recognition model training and abnormal behavior recognition method and device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106127525A (en) * | 2016-06-27 | 2016-11-16 | 浙江大学 | A kind of TV shopping Method of Commodity Recommendation based on sorting algorithm |
| CN106127546A (en) * | 2016-06-20 | 2016-11-16 | 重庆房慧科技有限公司 | A kind of Method of Commodity Recommendation based on the big data in intelligence community |
| CN107909433A (en) * | 2017-11-14 | 2018-04-13 | 重庆邮电大学 | A kind of Method of Commodity Recommendation based on big data mobile e-business |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
| US20180174188A1 (en) * | 2016-12-20 | 2018-06-21 | Wal-Mart Stores, Inc. | Systems and methods for customizing content of a billboard |
| US20180365715A1 (en) * | 2015-12-02 | 2018-12-20 | Tata Consultancy Services Limited | Method and system for purchase behavior prediction of customers |
| CN109741113A (en) * | 2019-01-10 | 2019-05-10 | 博拉网络股份有限公司 | A kind of user's purchase intention prediction technique based on big data |
| CN109741112A (en) * | 2019-01-10 | 2019-05-10 | 博拉网络股份有限公司 | A kind of user's purchase intention prediction technique based on mobile big data |
-
2019
- 2019-09-06 CN CN201910843151.6A patent/CN110544131A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180365715A1 (en) * | 2015-12-02 | 2018-12-20 | Tata Consultancy Services Limited | Method and system for purchase behavior prediction of customers |
| CN106127546A (en) * | 2016-06-20 | 2016-11-16 | 重庆房慧科技有限公司 | A kind of Method of Commodity Recommendation based on the big data in intelligence community |
| CN106127525A (en) * | 2016-06-27 | 2016-11-16 | 浙江大学 | A kind of TV shopping Method of Commodity Recommendation based on sorting algorithm |
| US20180174188A1 (en) * | 2016-12-20 | 2018-06-21 | Wal-Mart Stores, Inc. | Systems and methods for customizing content of a billboard |
| CN107909433A (en) * | 2017-11-14 | 2018-04-13 | 重庆邮电大学 | A kind of Method of Commodity Recommendation based on big data mobile e-business |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
| CN109741113A (en) * | 2019-01-10 | 2019-05-10 | 博拉网络股份有限公司 | A kind of user's purchase intention prediction technique based on big data |
| CN109741112A (en) * | 2019-01-10 | 2019-05-10 | 博拉网络股份有限公司 | A kind of user's purchase intention prediction technique based on mobile big data |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112270545A (en) * | 2020-10-27 | 2021-01-26 | 上海淇馥信息技术有限公司 | Financial risk prediction method and device based on migration sample screening and electronic equipment |
| CN113283960A (en) * | 2021-05-18 | 2021-08-20 | 武汉畅来电子商务有限公司 | Vertical e-commerce platform commodity intelligent recommendation method based on big data analysis and cloud computing and cloud service platform |
| CN113592160A (en) * | 2021-07-13 | 2021-11-02 | 北京淇瑀信息科技有限公司 | Fusion processing-based loss of contact risk prediction method and system for user equipment and computer equipment |
| CN114511022A (en) * | 2022-01-24 | 2022-05-17 | 百度在线网络技术（北京）有限公司 | Feature screening, behavior recognition model training and abnormal behavior recognition method and device |
| CN114511022B (en) * | 2022-01-24 | 2022-12-27 | 百度在线网络技术（北京）有限公司 | Feature screening, behavior recognition model training and abnormal behavior recognition method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110544131A (en) | Data-driven E-commerce user purchasing behavior prediction method | |
| US20180341898A1 (en) | Demand forecast | |
| CN106570718B (en) | Information delivery method and delivery system | |
| CN106991598A (en) | Data push method and its system | |
| US20030212619A1 (en) | Targeting customers | |
| CN110704730A (en) | Product data pushing method and system based on big data and computer equipment | |
| CN109636482B (en) | Data processing method and system based on similarity model | |
| CN110826886A (en) | Electric power customer portrait construction method based on clustering algorithm and principal component analysis | |
| US20150332292A1 (en) | System and method for monitoring market information for deregulated utilities based on transaction data | |
| CN110503446A (en) | The client segmentation method and decision-making technique of electric business platform based on clustering algorithm | |
| US20210109906A1 (en) | Clustering model analysis for big data environments | |
| CN113469730A (en) | Customer repurchase prediction method and device based on RF-LightGBM fusion model under non-contract scene | |
| US20220277331A1 (en) | Systems and methods for procurement cost forecasting | |
| CN107886241A (en) | Resource analysis method, apparatus, medium and electronic equipment | |
| CN111626767B (en) | Resource data issuing method, device and equipment | |
| JP6709775B2 (en) | Calculation device, calculation method, and calculation program | |
| CN112541111A (en) | Commodity retrieval and commodity recommendation system based on decentralized big data retrieval market | |
| CN115860880A (en) | Personalized commodity recommendation method and system based on multilayer heterogeneous graph convolution model | |
| CN115545886A (en) | Overdue risk identification method, overdue risk identification device, overdue risk identification equipment and storage medium | |
| CN111861679A (en) | Commodity recommendation method based on artificial intelligence | |
| CN110866698A (en) | Device for assessing service score of service provider | |
| CN111275371B (en) | Data processing method, data processing apparatus, and computer-readable storage medium | |
| CN116596576A (en) | Target recommendation method and device | |
| CN110490682B (en) | Method and device for analyzing commodity attributes | |
| US10346873B1 (en) | Method and system for implementing behavior isolating prediction model |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191206 |