CN110109901B - Method and device for screening target object - Google Patents
Method and device for screening target object Download PDFInfo
- Publication number
- CN110109901B CN110109901B CN201810105294.2A CN201810105294A CN110109901B CN 110109901 B CN110109901 B CN 110109901B CN 201810105294 A CN201810105294 A CN 201810105294A CN 110109901 B CN110109901 B CN 110109901B
- Authority
- CN
- China
- Prior art keywords
- data
- feature
- characteristic
- characteristic data
- product
- 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
- 238000012216 screening Methods 0.000 title claims abstract description 92
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012545 processing Methods 0.000 claims abstract description 87
- 238000004364 calculation method Methods 0.000 claims description 34
- 238000009825 accumulation Methods 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 19
- 230000002159 abnormal effect Effects 0.000 claims description 16
- 238000011156 evaluation Methods 0.000 claims description 16
- 230000000737 periodic effect Effects 0.000 claims description 16
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000009499 grossing Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 9
- 238000004140 cleaning Methods 0.000 claims description 8
- 230000006399 behavior Effects 0.000 description 80
- 238000010586 diagram Methods 0.000 description 19
- 230000006870 function Effects 0.000 description 13
- 238000004458 analytical method Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000002596 correlated effect Effects 0.000 description 6
- 230000003993 interaction Effects 0.000 description 5
- 230000009471 action Effects 0.000 description 4
- 238000010276 construction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000007635 classification algorithm Methods 0.000 description 2
- 238000005094 computer simulation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and a device for screening target objects, and relates to the technical field of computers. One embodiment of the method comprises the following steps: analyzing the initial data to generate characteristic data, and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data; calculating the association degree of the object and the product by utilizing the feature data after feature processing; and screening target objects meeting preset conditions according to the association degree. According to the method, the generated characteristic data are subjected to characteristic processing, and then the association degree of the object and the product is calculated, so that the target object meeting the requirements is screened, and the accuracy of screening the target object can be improved.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for screening a target object.
Background
In recent years, with continuous innovation and progress of information technology, ways of purchasing and selling goods, providing or receiving services and engaging in other business activities of users are also gradually and comprehensively internetworked, and how to find rules from user behavior data and predict future demands of users, so that recommending products to users in need is a key problem in accurate marketing of big data. Wherein, recommending the product to the user in need can be regarded as screening the target object according to the target product.
In the prior art method for screening target objects, conditions for screening target objects are manually defined according to historical experience of staff, for example, products are recommended to all objects purchased or browsed by the products.
In the process of implementing the present invention, the inventor finds that at least the following problems exist in the prior art: the method for screening the target object in the prior art is subjective, the target object is confirmed through the basic attribute and the behavior of the object, the accuracy is low, a good screening effect cannot be achieved, the efficiency is low, and the optimal time is easy to miss.
Disclosure of Invention
In view of this, the embodiment of the invention provides a method and a device for screening target objects, which perform feature processing on generated feature data, then calculate the association degree between the objects and products, and further screen target objects meeting the requirements, so as to improve the accuracy of screening target objects.
To achieve the above object, according to an aspect of an embodiment of the present invention, there is provided a method of screening a target object.
The method for screening the target object comprises the following steps: analyzing initial data to generate characteristic data, and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data; calculating the association degree of the object and the product by utilizing the feature data after feature processing; and screening target objects meeting preset conditions according to the association degree.
Optionally, before analyzing the initial data to generate the feature data, the method further comprises: obtaining initial data, wherein the object information data comprises: object behavior data and object base data, the product information data including product evaluation data and product base data; and cleaning the initial data to remove abnormal data in the initial data.
Optionally, analyzing the initial data to generate the feature data includes: and combining the initial data according to different dimensions to establish the characteristic data.
Optionally, performing feature processing on the feature data includes: performing attenuation processing on the characteristic data based on the time accumulation attribute; and/or smoothing the feature data based on the attribute negatively correlated with the degree of correlation; and/or discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing.
Optionally, the feature data after feature processing includes: first characteristic data and second characteristic data, wherein the first characteristic data and the second characteristic data are data of the same structure, but the data time of the first characteristic data is earlier than the data time of the second characteristic data.
Optionally, calculating the association degree of the object and the product by using the feature data after the feature processing includes: training a computing model by using the first characteristic data, wherein the input of the computing model is characteristic data, and the output is the association degree of the object and the product; and calculating the association degree of the object and the product in the second characteristic data by using the calculation model.
Optionally, screening the target object meeting the preset condition according to the association degree includes: and screening target objects with the association degree with the target products being larger than a preset threshold according to screening requirements of the target products, wherein the target products are products with screening requirements in the second characteristic data.
To achieve the above object, according to another aspect of the embodiments of the present invention, there is provided an apparatus for screening a target object.
The device for screening the target object comprises: the processing module is used for analyzing the initial data to generate characteristic data and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data; the computing module is used for computing the association degree of the object and the product by utilizing the feature data after the feature processing; and the screening module is used for screening the target objects meeting preset conditions according to the association degree.
Optionally, the apparatus further comprises: the acquisition module is used for acquiring initial data, wherein the object information data comprises: object behavior data and object basic data, wherein the product information data comprises product evaluation data and product basic data; and cleaning the initial data to remove abnormal data in the initial data.
Optionally, the processing module is further configured to: and combining the initial data according to different dimensions to establish the characteristic data.
Optionally, the processing module is further configured to: performing attenuation processing on the characteristic data based on the time accumulation attribute; and/or smoothing the feature data based on the attribute negatively correlated with the degree of correlation; and/or discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing.
Optionally, the feature data after feature processing includes: first characteristic data and second characteristic data, wherein the first characteristic data and the second characteristic data are data of the same structure, but the data time of the first characteristic data is earlier than the data time of the second characteristic data.
Optionally, the computing module is further configured to: training a computing model by using the first characteristic data, wherein the input of the computing model is characteristic data, and the output is the association degree of the object and the product; and calculating the association degree of the object and the product in the second characteristic data by using the calculation model.
Optionally, the screening module is further configured to: and screening target objects with the association degree with the target products being larger than a preset threshold according to screening requirements of the target products, wherein the target products are products with screening requirements in the second characteristic data.
To achieve the above object, according to still another aspect of the embodiments of the present invention, there is provided an electronic device.
An electronic device of an embodiment of the present invention includes: one or more processors; and the storage device is used for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors are enabled to realize the method for screening the target object according to the embodiment of the invention.
To achieve the above object, according to still another aspect of the embodiments of the present invention, there is provided a computer-readable medium.
A computer readable medium of an embodiment of the present invention has stored thereon a computer program which, when executed by a processor, implements a method of screening a target object of the embodiment of the present invention.
One embodiment of the above invention has the following advantages or benefits: performing feature processing on the generated feature data, then calculating the association degree of the object and the product, and further screening out target objects meeting the requirements according to the association degree, so that the accuracy of screening the target objects can be improved; in the embodiment of the invention, the acquired initial data is cleaned, so that abnormal data can be removed, and the accuracy of an algorithm model is improved; according to the embodiment of the invention, the initial data are combined to construct the characteristic data according to different dimensions, so that the diversification of the characteristic data can be ensured; in the embodiment of the invention, the characteristic data is subjected to characteristic processing from a plurality of angles such as time accumulation attribute, correlation degree negative correlation attribute, periodic attribute and the like of the characteristic data, so that the accuracy of the characteristic data for training a calculation model can be further improved; the feature data after feature processing in the embodiment of the invention comprises first feature data and second feature data, which have the same structure and different time, so that a calculation model can be trained by using the first feature data, and the association degree between an object and a product in the second feature data is calculated by using the trained calculation model; according to the method and the device for selecting the target objects, the target objects meeting the preset conditions are selected by utilizing the association degree according to the screening requirements of the target products, so that the preset conditions can be set according to specific service requirements, and the target objects meeting the preset conditions can be selected.
Further effects of the above-described non-conventional alternatives are described below in connection with the embodiments.
Drawings
The drawings are included to provide a better understanding of the invention and are not to be construed as unduly limiting the invention. Wherein:
FIG. 1 is a schematic diagram of the main steps of a method of screening a target object according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the main architecture of a system suitable for implementing the method of screening a target object of an embodiment of the present invention;
FIG. 3 is a schematic graph of an attenuation function of a method of screening a target object according to an embodiment of the present invention;
FIGS. 4 and 5 are schematic diagrams of a method of screening a target object according to an embodiment of the present invention;
FIG. 6 is a schematic diagram of the main flow of a method of screening a target object according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of main modules of an apparatus for screening a target object according to an embodiment of the present invention;
FIG. 8 is an exemplary system architecture diagram in which embodiments of the present invention may be applied;
fig. 9 is a schematic diagram of a computer system suitable for use in implementing an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, in which various details of the embodiments of the present invention are included to facilitate understanding, and are to be considered merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
Fig. 1 is a schematic diagram of main steps of a method for screening a target object according to an embodiment of the present invention, as shown in fig. 1, the method for screening a target object according to an embodiment of the present invention mainly includes the following steps:
Step S101: analyzing the initial data to generate characteristic data, and carrying out characteristic processing on the characteristic data. Wherein the initial data may include object information data and product information data. The invention aims to screen out target objects meeting the requirements according to target products. Accordingly, the initial data employed includes object information data and product information data. In the invention, in order to improve the accuracy of the screened target object, the generated characteristic data is subjected to characteristic processing.
Step S102: and calculating the association degree of the object and the product by using the feature data after the feature processing. In the invention, the association degree of the object and the product can be adjusted according to the actual scene, for example, in the scene that the e-commerce platform predicts the purchasing behavior of the user, the association degree of the object and the product can be the possibility that the user purchases a certain product.
Step S103: and screening target objects meeting preset conditions according to the association degree. In the embodiment of the invention, the target object meeting the preset condition is screened by utilizing the association degree calculated in the step S102, so that the preset condition can be set according to the specific service requirement, and the target object meeting the preset condition can be selected.
In the embodiment of the present invention, before analyzing the initial data to generate the feature data, the method for screening the target object may further include: acquiring initial data; and cleaning the initial data to remove abnormal data in the initial data. Wherein the object information data may include: the object behavior data and the object base data, and the product information data may include product evaluation data and product base data. The object behavior data includes behavior time data and behavior classification data. The behavior classification includes the behavior of object clicking, browsing, focusing, purchasing, and the like. The object basic data is basic attribute information data of the object, such as information of a name, a category, a registration date, and a level. The product evaluation data refers to evaluation information of the object on the product, such as a poor evaluation rate, a good evaluation rate, an accumulated comment number, and the like. The product basic data is basic information data of the product, such as information of category, brand, price, attribute and the like. In the embodiment of the invention, the acquired initial data is cleaned, and the abnormal data is removed, wherein the abnormal data can comprise: redundancy recorded data, data generated by a crawler user, data with abnormal attribute information, and the like. The redundant record is a redundant record generated when data is acquired by using a technical means. The crawler user refers to a 'web robot', and is a behavior for automatically completing the operation on the website through programming. The attribute information abnormality refers to information abnormality of a product or information abnormality of an object, such as an excessively high price of the product, a registration date of the object being 100 years before the male element, and the like. According to the embodiment of the invention, the acquired initial data is cleaned, so that abnormal data can be removed, and the accuracy of the algorithm model is improved.
In the embodiment of the present invention, the analyzing the initial data to generate the feature data may include: and combining the initial data according to the different dimensions to establish characteristic data. According to the embodiment of the invention, the initial data are combined to construct the characteristic data according to different dimensions, so that the diversification of the characteristic data can be ensured.
In the embodiment of the present invention, performing feature processing on feature data may include: performing attenuation processing on the characteristic data based on the time accumulation attribute; and/or smoothing the feature data based on the attribute negatively correlated with the degree of correlation; and/or discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing. In the invention, based on the time accumulation attribute, the characteristic data can be divided into time accumulation characteristic data and non-time accumulation characteristic data, wherein the time accumulation characteristic data is related to the accumulation amount of the object for multiple days, and the time accumulation characteristic data has different weights due to different lengths of the time accumulation characteristic data from the current time, so that the attenuation processing is carried out on the characteristics. The attribute of negative correlation with the correlation degree means that when the accumulated value of some feature data is too high, negative correlation influence is generated on the correlation degree, so that smoothing processing is required to be carried out on peak parts of the feature data, and the peak is eliminated. The discretization process in the invention is convenient for explaining and outlining the actual object behavior data. The periodic attribute refers to the periodic behavior of the object, and the operation performed in a period of time is called one-time behavior, and feature data belonging to the same time are combined. In the embodiment of the invention, the characteristic data is subjected to characteristic processing from a plurality of angles such as time accumulation attribute, correlation degree negative correlation attribute, periodic attribute and the like of the characteristic data, so that the accuracy of the characteristic data for training a calculation model can be further improved.
In the embodiment of the present invention, the feature data after feature processing may include: first feature data and second feature data. Calculating the association degree of the object and the product using the feature data after the feature processing may include: training a computing model by using the first characteristic data, wherein the input of the trained computing model is the characteristic data, and the output is the association degree of the object and the product; and calculating the association degree of the object and the product in the second characteristic data by using the calculation model. In the invention, the first characteristic data and the second characteristic data are data with the same structure, the first characteristic data are used for training a calculation model, and the second characteristic data are used for predicting the behavior of the object, so that the data time of the second characteristic data is closer to the predicted date, that is, the data time of the first characteristic data is earlier than the data time of the second characteristic data.
In the embodiment of the present invention, according to the association degree, the screening of the target object meeting the preset condition may include: and screening target objects with the association degree with the target products larger than a preset threshold according to the screening requirements of the target products. The target product is a product with screening requirements in the second characteristic data. In the invention, the preset conditions can be the screening of a preset number of target objects, or the screening of target objects with the association degree with the target product being larger than a certain threshold, and of course, the invention does not need to limit the target objects too much, and the preset conditions can be flexibly defined according to actual requirements.
In order to facilitate understanding, in the following description of the present invention, taking an example that an e-commerce platform predicts the purchasing behavior of a user, taking a product as a commodity, taking an object as a user, taking product information data as commodity information data, taking object information data as user information data, taking product evaluation data as commodity evaluation data, taking product basic data as commodity basic data, taking object behavior data as user behavior data, taking object basic data as user basic data, taking first characteristic data as a characteristic data table of a construction model, taking second characteristic data as a characteristic table of predicted purchasing intent, taking association degree of an object and a product as intent value of purchasing commodity by the user, of course, "commodity," "user," "commodity information data," "user information data," "commodity evaluation data," "commodity basic data," "user behavior data," "user basic data," "feature table of construction model," "feature table of predicted purchase," and "intention value of user to purchase commodity" are not intended to limit the scope of protection of the technical solution of the present invention, and in the present invention, "product," "object," "product information data," "object information data," "product evaluation data," "product basic data," "object behavior data," "object basic data," "first feature data," "second feature data," and "association degree of object and product" may be adaptively adjusted according to specific service scenarios. The intention value of a user to purchase a commodity refers to the likelihood that the user purchases the commodity, and the intention value is higher as the likelihood that the user purchases a certain commodity is higher.
In the embodiment of the invention, the method for screening the target object can be specifically realized by designing a system for screening the target object. Fig. 2 is a schematic diagram of the main architecture of a system suitable for implementing the method of screening a target object according to an embodiment of the present invention. As shown in fig. 2, a system suitable for implementing a method for screening a target object according to an embodiment of the present invention may include: the system comprises a data access unit, a behavior analysis unit, a purchase prediction unit and a data output unit. The data access unit is used for acquiring data from the big data warehouse and cleaning the acquired data. The behavior analysis unit is used for establishing a characteristic table of the multi-dimensional construction model according to the combination of the user behavior data and the basic data, and constructing an algorithm model according to the characteristic data. The purchase prediction unit is used for calculating possible values of the commodity purchased by the user according to the constructed algorithm model. The data output unit is used for outputting users who are likely to purchase the commodities in the selected commodity set.
In the data access unit, the acquired data may include: user base data, commodity evaluation data, and user behavior data. User base data, mainly from user registration and consumption statistics of e-commerce platform websites, may include: age, gender, user rating, registration date, etc. The commodity basic data mainly comes from commodity acquisition and marketing data of an electronic commerce platform, and can comprise: properties of the merchandise, category, brand, etc. The commodity evaluation data mainly comes from comment and scoring summarized data of purchased commodities by users, and can comprise: whether there is a bad comment, a bad comment rate, an accumulated comment number, and the like. The user behavior data may include: the time it takes for a user to click, browse, pay attention to, join a shopping cart, delete a shopping cart, purchase, etc., to perform these actions, i.e., the action time. The user clicks and browses the flow data generated after the data comes from the front end buried point (buried point means that user behavior is collected by using a technical means) of the e-commerce platform website, and the sources for collecting the flow data can include: WEB sites, mobile APP (Application), mobile WAP (Wireless Application Protocol) sites, third party portals, etc. Other user behavior data comes from the user's operational data on the merchandise in the e-commerce platform website.
In the embodiment of the invention, the data can be acquired through task scheduling of the large data warehouse. Wherein the data for constructing the calculation model and the data for predicting the purchase intention are taken from data of different time periods, and the time period of the data for predicting the purchase intention is closer to the time for promotion, for example, the date of commodity promotion is 11 th day, the data of 1 st day to 5 th day may be selected for constructing the calculation model, and the data of 6 th day to 10 th day may be selected for predicting the purchase intention.
After the data access unit acquires data from the large data warehouse, the acquired data is cleaned. Wherein, the cleaning of the data may include: removing repeated clicking and browsing data; removing the influence of the behavior of the crawler user on the data; abnormal values of the washing data, for example, the age of the user is 1000 years, the commodity price is excessively high, and the like.
And the behavior analysis unit combines the user behavior data and the basic data according to different dimensions to generate feature data. And then, using Pearson correlation coefficient (Pearson Correlation Coefficient, which is used for measuring whether two data sets are on a line or not and used for measuring the linear relation between distance variables) to measure the linear relation between the generated characteristic data and the purchasing behavior, and selecting the characteristic data which are positively correlated and related to the purchasing behavior as a characteristic table for constructing a model.
Taking the e-commerce platform to predict the purchasing behavior of the user as an example, the invention can be used for constructing the characteristics in the characteristic table of the model, and the characteristics can comprise: the user average behavior period (the user average behavior period refers to a time period for a user to perform a behavior operation), the time when the user clicks a commodity, the time when the user browses the commodity, the time when the commodity is purchased, the latest time interval of the user to the class (the class refers to the class to which the commodity belongs), the time when the user browses the commodity, the time when the user acts on the commodity (the time when the behavior is used for the user to perform clicking, browsing, focusing, adding a shopping cart, deleting a shopping cart, purchasing and the like), the behavior number within seven days of the user (the behavior number refers to the number of times when the user performs clicking, browsing, focusing, adding a shopping cart, deleting a shopping cart, purchasing and the like), the behavior number within the last time of the user in the behavior period, the user's active day (the user is considered to be active for the day when the behavior of the commodity exceeds a certain time), the user clicks the average time when the commodity is used, the purchase number of the user purchases the commodity, the browse number of the commodity, and the purchase rate.
In addition to the important feature data described above, the following secondary features are also important indicators for constructing a computational model: the number of days of behavior in a characteristic day interval (the characteristic day interval can be a promotion time interval, a holiday time interval and the like), the number of hours of a predicted day of a first/last behavior distance of a user, the average value of each commodity behavior of the user, the total number of active historic days of the user, the number of continuous login days of the user, the average number of commodities operated daily by the user, the ordering of active days of the user with the same first interaction time (the interaction time refers to the fact that more than one user performs behavior operation at a certain time and is considered to be the interaction time), the duration of the last three active periods, the number of interaction time periods, the sum of the median of the maximum and minimum average values of the interaction time intervals, the specific gravity of the commodity of the current class, the average number of active minutes daily of the user, the center of active time of the user, the difference between the last browsing time of the user and the first browsing time of the user, the growth speed of the user class and the like are seen in the user login behavior.
In the behavior analysis unit, feature processing is performed on a feature table used for constructing the calculation model. Wherein, the feature processing may include: and carrying out attenuation processing on the feature table constructing the calculation model based on the time accumulation attribute. In the present invention, the attenuation process may be to add an attenuation function to the time-integrated characteristic data. The formula of the attenuation function may be: D is the time difference of day of statistical prediction of distance, and gamma is the attenuation coefficient (the attenuation coefficient gamma is obtained empirically and takes on a value between 0.3 and 0.95). Fig. 3 is a schematic diagram of an attenuation function curve of a method of screening a target object according to an embodiment of the present invention. Fig. 3 shows a graph of the parameter distribution of the daily decay function at gamma of 0.7. Taking the user click data as an example, the statistical method of adding the user click data into the decay function is as follows:
Sum＝d1*1.0+d2*0.78+d3*0.562+d4*0.438+…
where di is the user click amount on the i-th day from the commodity sales date, sum is the total value of the click amount in the user statistics time, i.e. the final user click feature data. In the embodiment of the invention, the Pearson correlation coefficient of the single feature and the purchasing behavior after the attenuation function is added can be improved by more than 30%.
At the behavior analysis unit, the feature processing may include: and smoothing the feature table constructing the calculation model based on the attribute negatively correlated with the association degree. For example, in the embodiment of the invention, the total value of the behaviors such as clicking and browsing by a user has the data value of the behaviors which are clicked and browsed by a robot but not purchased, so that the smoothing processing is carried out on the behaviors, and the influence of the behaviors is reduced.
At the behavior analysis unit, the feature processing may include: and carrying out discretization processing on the feature table constructing the calculation model based on the periodic attribute, and merging feature data belonging to the same period after the discretization processing. The discretization process in the invention is convenient for explaining and outlining the actual user behavior data. The periodic attribute refers to the periodic behavior of the user, and the operation performed in a period of time is called a behavior, for example, a certain user browses some commodities between 18-20 yesterday late and 8-11 days browses some commodities, so that the operation is called two behavior periods. In the present invention, operations with an operation time difference of less than 45 minutes (45 minutes is generally considered as a website login failure time, or other time, which is not limited in the present invention), which are two operations before and after the user, may be combined together, and a behavior exceeding 45 minutes is considered as a new user behavior period. Fig. 4 and 5 are schematic diagrams of a method of screening a target object according to an embodiment of the present invention. Fig. 4 shows clicking behaviors of a user in a day, and the behaviors of the user shown in fig. 4 are combined into two behavior periods, wherein each period contains a certain number of user behaviors, so as to obtain a schematic diagram after the combination processing shown in fig. 5. In the embodiment of the invention, the accuracy of the algorithm model can be effectively improved by counting the characteristic data such as the times of the user behavior period, the times of the behaviors in the recent behavior period, the activity of the behavior period and the like.
After the behavior analysis unit performs feature processing on the feature table for constructing the calculation model, the calculation model may be trained using XGBOOST (an algorithm tool for user machine learning) two-classification algorithm. Empirically, XGBOOST parameters were as follows:
'learning_rate':0.1,'n_estimators':1000,
'max_depth':8,'min_child_weight':3,
'subsample':1.0,'colsample_bytree':0.8,
'eta':0.1,'objective':'binary:logistic',
'gamma':0.1,'lambda':550.
When the calculation model is trained, a model of whether a user purchases or not can be built first, and then a model of whether the user purchases the appointed commodity or not is built. Of course, other two-class machine learning algorithms, such as SVM (support vector machine, a trainable machine learning method) and the like, can also be used in the present invention to construct the computational model. And (3) performing model training on the feature table of the constructed model, and storing the training result in a model file.
In the purchase prediction unit, the structure of the adopted prediction data is the same as that of the data of the algorithm model constructed by the behavior analysis unit, but the time is different. And calling the same feature processing method by the predicted data to generate a feature table of the predicted purchase intention, inputting the feature table of the predicted purchase intention into a model file generated in the behavior analysis unit to calculate, and outputting the purchase intention value and ranking of the commodity by the user.
In the data output unit, the user group that may purchase the commodity set in the estimated time may be selected by the commodity set provided by the operator, and the user group that purchases the commodity set may be selected as long as the commodity in the set may be purchased by a user in the user group. The method for selecting the user group in the invention can judge through the threshold value, for example, the commodity is recommended to the user with the intention value larger than the preset threshold value, or can select the preset number of users, for example, the commodity in the commodity set is recommended to 100 users in a preset manner, and then the commodity is recommended to the user with the intention value ranked at the top 100.
Fig. 6 is a schematic diagram of a main flow of a method of screening a target object according to an embodiment of the present invention. Taking an e-commerce platform as an example, the basic data is stable, the user behavior data of 6-10 days before the predicted purchase day is selected to be used for constructing an algorithm model, the user behavior data of 1-5 days before the predicted purchase day is selected to be used for predicting the purchase intention, and of course, the invention can also select data of other time periods, and only the data used for constructing the algorithm model and the data used for predicting the purchase intention are required to be data which are not crossed in time.
As shown in fig. 6, the main flow of the method for screening a target object may include: step S601, basic data and user behavior data are obtained, the obtained data are cleaned, and abnormal data are removed; step S602, combining the basic data and user behavior data of 6-10 days before predicted purchase according to different dimensions to establish a multi-dimensional feature table for constructing a model; step S603, performing attenuation processing on the characteristics related to the accumulated number of the user in the characteristic table; step S604, smoothing the characteristics related to the clicking and browsing actions of the user in the characteristic table; step S605, combining the characteristics related to the user behavior period in the characteristic table; step S606, training a calculation model by using the processed feature table according to the two classification algorithm of XGBOOST; step S607, according to different dimensions, combining the basic data and the user behavior data 1-5 days before the predicted purchase day to establish a multi-dimensional feature table for predicting the purchase intention, and calling the same feature processing method to perform feature processing on the feature table; step S608, inputting a feature table for predicting purchase intention into a calculation model to obtain purchase intention value and ranking of the corresponding product of the user; step S609, selecting a user who is likely to purchase the commodity set in the predicted purchase date according to the commodity set provided by the operator.
In the present invention, the steps S603, S604 and S605 are all feature processing on the feature table, and the execution sequence of these feature processing may be adjusted according to the actual situation, which is not limited in the present invention.
According to the technical scheme for screening the target object, disclosed by the embodiment of the invention, the generated characteristic data is subjected to characteristic processing, and then the association degree of the object and a product is calculated, so that the target object meeting the requirements is screened, and the accuracy of screening the target object can be improved; in the embodiment of the invention, the acquired initial data is cleaned, so that abnormal data can be removed, and the accuracy of an algorithm model is improved; according to the embodiment of the invention, the initial data are combined to construct the characteristic data according to different dimensions, so that the diversification of the characteristic data can be ensured; in the embodiment of the invention, the characteristic data is subjected to characteristic processing from a plurality of angles such as time accumulation attribute, correlation degree negative correlation attribute, periodic attribute and the like of the characteristic data, so that the accuracy of the characteristic data for training a calculation model can be further improved; the feature data after feature processing in the embodiment of the invention comprises first feature data and second feature data, which have the same structure and different time, so that a calculation model can be trained by using the first feature data, and the association degree between an object and a product in the second feature data is calculated by using the trained calculation model; according to the method and the device for selecting the target objects, the target objects meeting the preset conditions are selected by utilizing the association degree according to the screening requirements of the target products, so that the preset conditions can be set according to specific service requirements, and the target objects meeting the preset conditions can be selected.
Fig. 7 is a schematic diagram of main modules of an apparatus for screening a target object according to an embodiment of the present invention. As shown in fig. 7, an apparatus 700 for screening a target object according to an embodiment of the present invention mainly includes the following modules: a processing module 701, a computing module 702 and a screening module 703.
The processing module 701 may be configured to analyze the initial data to generate feature data, and perform feature processing on the feature data. The initial data may include: object information data and product information data. The computing module 702 may be configured to calculate a degree of association between the object and the product using the feature data after the feature processing. The screening module 703 may be configured to screen the target object according to the association degree.
In an embodiment of the present invention, the apparatus 700 for screening a target object may further include: an acquisition module (shown in the figure). The acquisition module can be used for acquiring initial data; and cleaning the initial data to remove abnormal data in the initial data. Wherein the object information data may include: object behavior data and object base data, the product information data may include: product evaluation data and product base data.
In an embodiment of the present invention, the processing module 701 may further be configured to: and combining the initial data according to the different dimensions to establish characteristic data.
In an embodiment of the present invention, the processing module 701 may further be configured to: performing attenuation processing on the characteristic data based on the time accumulation attribute; and/or smoothing the feature data based on the attribute negatively correlated with the degree of correlation; and/or discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing.
In the embodiment of the present invention, the feature data after feature processing may include: first feature data and second feature data. The first characteristic data and the second characteristic data are data with the same structure, but the data time of the first characteristic data is earlier than the data time of the second characteristic data.
In an embodiment of the present invention, the computing module 702 may be further configured to: training a computing model using the first feature data; and calculating the association degree of the object and the product in the second characteristic data by using the calculation model. The input of the calculation model is characteristic data, and the output is the association degree of the object and the product.
In the embodiment of the present invention, the screening module 703 may further be used for: and screening target objects with the association degree with the target products being greater than a preset threshold according to the screening requirements of the target products. The target product is a product with screening requirements in the second characteristic data.
From the above description, it can be seen that the generated feature data is subjected to feature processing, and then the association degree of the object and the product is calculated, so as to screen the target object meeting the requirements, thereby improving the accuracy of screening the target object; in the embodiment of the invention, the acquired initial data is cleaned, so that abnormal data can be removed, and the accuracy of an algorithm model is improved; according to the embodiment of the invention, the initial data are combined to construct the characteristic data according to different dimensions, so that the diversification of the characteristic data can be ensured; in the embodiment of the invention, the characteristic data is subjected to characteristic processing from a plurality of angles such as time accumulation attribute, correlation degree negative correlation attribute, periodic attribute and the like of the characteristic data, so that the accuracy of the characteristic data for training a calculation model can be further improved; the feature data after feature processing in the embodiment of the invention comprises first feature data and second feature data, which have the same structure and different time, so that a calculation model can be trained by using the first feature data, and the association degree between an object and a product in the second feature data is calculated by using the trained calculation model; according to the method and the device for selecting the target objects, the target objects meeting the preset conditions are selected by utilizing the association degree according to the screening requirements of the target products, so that the preset conditions can be set according to specific service requirements, and the target objects meeting the preset conditions can be selected.
Fig. 8 illustrates an exemplary system architecture 800 of a method of screening a target object or an apparatus of screening a target object to which embodiments of the present invention may be applied.
As shown in fig. 8, a system architecture 800 may include terminal devices 801, 802, 803, a network 804, and a server 805. The network 804 serves as a medium for providing communication links between the terminal devices 801, 802, 803 and the server 805. The network 804 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.
A user may interact with the server 805 through the network 804 using the terminal devices 801, 802, 803 to receive or send messages or the like. Various communication client applications such as shopping class applications, web browser applications, search class applications, instant messaging tools, mailbox clients, social platform software, etc. (by way of example only) may be installed on the terminal devices 801, 802, 803.
The terminal devices 801, 802, 803 may be a variety of electronic devices having a display screen and supporting web browsing, including but not limited to smartphones, tablets, laptop and desktop computers, and the like.
The server 805 may be a server providing various services, such as a background management server (by way of example only) that provides support for shopping-type websites browsed by users using the terminal devices 801, 802, 803. The background management server may analyze and process the received data such as the product information query request, and feedback the processing result (e.g., the target push information, the product information—only an example) to the terminal device.
It should be noted that, the method for screening a target object according to the embodiment of the present invention is generally executed by the server 805, and accordingly, the device for screening a target object is generally disposed in the server 805.
It should be understood that the number of terminal devices, networks and servers in fig. 8 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
Referring now to FIG. 9, there is illustrated a schematic diagram of a computer system 900 suitable for use in implementing an embodiment of the present invention. The terminal device shown in fig. 9 is only an example, and should not impose any limitation on the functions and the scope of use of the embodiment of the present invention.
As shown in fig. 9, the computer system 900 includes a Central Processing Unit (CPU) 901, which can execute various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 902 or a program loaded from a storage section 908 into a Random Access Memory (RAM) 903. In the RAM 903, various programs and data necessary for the operation of the system 900 are also stored. The CPU 901, ROM 902, and RAM 903 are connected to each other through a bus 904. An input/output (I/O) interface 905 is also connected to the bus 904.
The following components are connected to the I/O interface 905: an input section 906 including a keyboard, a mouse, and the like; an output portion 907 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and a speaker; a storage portion 908 including a hard disk or the like; and a communication section 909 including a network interface card such as a LAN card, a modem, or the like. The communication section 909 performs communication processing via a network such as the internet. The drive 910 is also connected to the I/O interface 905 as needed. A removable medium 911 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed as needed on the drive 910 so that a computer program read out therefrom is installed into the storage section 908 as needed.
In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flow chart. In such an embodiment, the computer program may be downloaded and installed from the network via the communication portion 909 and/or installed from the removable medium 911. The above-described functions defined in the system of the present invention are performed when the computer program is executed by a Central Processing Unit (CPU) 901.
The computer readable medium shown in the present invention may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules involved in the embodiments of the present invention may be implemented in software or in hardware. The described modules may also be provided in a processor, for example, as: a processor includes a processing module, a computing module, and a screening module. The names of these modules do not limit the module itself in some cases, and for example, the processing module may also be described as "a module that analyzes initial data to generate feature data and performs feature processing on the feature data".
As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be present alone without being fitted into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to include: analyzing the initial data to generate characteristic data, and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data; calculating the association degree of the object and the product by utilizing the feature data after feature processing; and screening target objects meeting preset conditions according to the association degree.
According to the technical scheme of the embodiment of the invention, the generated characteristic data is subjected to characteristic processing, and then the association degree of the object and the product is calculated, so that the target object meeting the requirements is screened, and the accuracy of screening the target object can be improved; in the embodiment of the invention, the acquired initial data is cleaned, so that abnormal data can be removed, and the accuracy of an algorithm model is improved; according to the embodiment of the invention, the initial data are combined to construct the characteristic data according to different dimensions, so that the diversification of the characteristic data can be ensured; in the embodiment of the invention, the characteristic data is subjected to characteristic processing from a plurality of angles such as time accumulation attribute, correlation degree negative correlation attribute, periodic attribute and the like of the characteristic data, so that the accuracy of the characteristic data for training a calculation model can be further improved; the feature data after feature processing in the embodiment of the invention comprises first feature data and second feature data, which have the same structure and different time, so that a calculation model can be trained by using the first feature data, and the association degree between an object and a product in the second feature data is calculated by using the trained calculation model; according to the method and the device for selecting the target objects, the target objects meeting the preset conditions are selected by utilizing the association degree according to the screening requirements of the target products, so that the preset conditions can be set according to specific service requirements, and the target objects meeting the preset conditions can be selected.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives can occur depending upon design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method of screening a target object, comprising:
analyzing initial data to generate characteristic data, and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data;
Calculating the association degree of the object and the product by utilizing the feature data after feature processing; the feature data after feature processing includes: first and second feature data;
screening target objects meeting preset conditions according to the association degree;
Wherein, the performing feature processing on the feature data includes:
Performing attenuation processing on the characteristic data based on the time accumulation attribute; the attenuation processing comprises the step of adding an attenuation function into characteristic data of time accumulation attributes, wherein the attenuation function formula is as follows: d is the time difference of the distance statistics prediction day, and gamma is the attenuation coefficient; and
Smoothing the characteristic data based on the attribute negatively related to the association degree; and
Discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing;
calculating the association degree of the object and the product by using the feature data after feature processing comprises the following steps:
training a computing model by using the first characteristic data, wherein the input of the computing model is characteristic data, and the output is the association degree of the object and the product; training a calculation model, namely establishing whether an object purchases a model and establishing whether the object purchases a specified product model, and training by adopting a two-class algorithm;
Calculating the association degree of the object and the product in the second characteristic data by using the calculation model;
the analyzing the initial data to generate the characteristic data comprises the following steps: combining the initial data according to different dimensions to establish the characteristic data;
after analyzing the initial data to generate the characteristic data, the method further comprises: acquiring the linear relation between the characteristic data and the purchasing behavior by adopting a Pearson correlation coefficient; and selecting the feature data which are positively related and related to the purchasing behavior based on the linear relation to perform feature processing.
2. The method of claim 1, wherein prior to analyzing the initial data to generate the feature data, the method further comprises:
obtaining initial data, wherein the object information data comprises: object behavior data and object base data, the product information data including product evaluation data and product base data;
And cleaning the initial data to remove abnormal data in the initial data.
3. The method of claim 1, wherein the first feature data and the second feature data are data of the same structure, but the first feature data has a data time earlier than the second feature data.
4. The method of claim 1, wherein screening for target objects meeting a preset condition according to the degree of association comprises:
and screening target objects with the association degree with the target products being larger than a preset threshold according to screening requirements of the target products, wherein the target products are products with screening requirements in the second characteristic data.
5. An apparatus for screening a target object, comprising:
the processing module is used for analyzing the initial data to generate characteristic data and carrying out characteristic processing on the characteristic data, wherein the initial data comprises object information data and product information data;
The computing module is used for computing the association degree of the object and the product by utilizing the feature data after the feature processing; the feature data after feature processing includes: first and second feature data;
The screening module is used for screening target objects meeting preset conditions according to the association degree;
Wherein the processing module is further configured to:
Performing attenuation processing on the characteristic data based on the time accumulation attribute; the attenuation processing comprises the step of adding an attenuation function into characteristic data of time accumulation attributes, wherein the attenuation function formula is as follows: d is the time difference of the distance statistics prediction day, and gamma is the attenuation coefficient; and
Smoothing the characteristic data based on the attribute negatively related to the association degree; and
Discretizing the characteristic data based on the periodic attribute, and merging the characteristic data belonging to the same period after the discretizing;
The computing module is further for:
training a computing model by using the first characteristic data, wherein the input of the computing model is characteristic data, and the output is the association degree of the object and the product; training a calculation model, namely establishing whether an object purchases a model and establishing whether the object purchases a specified product model, and training by adopting a two-class algorithm;
Calculating the association degree of the object and the product in the second characteristic data by using the calculation model;
The processing module is further configured to: combining the initial data according to different dimensions to establish the characteristic data; acquiring the linear relation between the characteristic data and the purchasing behavior by adopting a Pearson correlation coefficient; and selecting the feature data which are positively related and related to the purchasing behavior based on the linear relation to perform feature processing.
6. The apparatus of claim 5, wherein the apparatus further comprises: the acquisition module is used for acquiring initial data, wherein the object information data comprises: object behavior data and object basic data, wherein the product information data comprises product evaluation data and product basic data; and cleaning the initial data to remove abnormal data in the initial data.
7. The apparatus of claim 5, wherein the first characteristic data and the second characteristic data are data of a same structure, but the first characteristic data has a data time earlier than the second characteristic data.
8. The apparatus of claim 5, wherein the screening module is further configured to:
and screening target objects with the association degree with the target products being larger than a preset threshold according to screening requirements of the target products, wherein the target products are products with screening requirements in the second characteristic data.
9. An electronic device, comprising:
One or more processors;
Storage means for storing one or more programs,
When executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-4.
10. A computer readable medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810105294.2A CN110109901B (en) | 2018-02-02 | 2018-02-02 | Method and device for screening target object |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810105294.2A CN110109901B (en) | 2018-02-02 | 2018-02-02 | Method and device for screening target object |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110109901A CN110109901A (en) | 2019-08-09 |
| CN110109901B true CN110109901B (en) | 2024-05-24 |
Family
ID=67483554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810105294.2A Active CN110109901B (en) | 2018-02-02 | 2018-02-02 | Method and device for screening target object |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110109901B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112488840B (en) * | 2019-09-12 | 2024-08-16 | 京东科技控股股份有限公司 | Information output method and device |
| CN110764726B (en) * | 2019-10-18 | 2023-08-22 | 网易（杭州）网络有限公司 | Target object determination method and device, terminal equipment and storage medium |
| CN111241081A (en) * | 2020-01-09 | 2020-06-05 | 杭州涂鸦信息技术有限公司 | IOT platform data collection method and system, readable storage medium and computer equipment |
| CN111369324B (en) * | 2020-03-12 | 2024-01-23 | 苏州大学 | Target information determining method, device, equipment and readable storage medium |
| CN112015726B (en) * | 2020-08-21 | 2024-04-12 | 广州欢网科技有限责任公司 | User activity prediction method, system and readable storage medium |
| CN113076346A (en) * | 2021-03-31 | 2021-07-06 | 北京深演智能科技股份有限公司 | Method and device for determining target object |
| CN118568143A (en) * | 2023-04-11 | 2024-08-30 | 上海御灵树网络科技有限公司 | Big data analysis method with screening function |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102208087A (en) * | 2010-03-30 | 2011-10-05 | 株式会社日立制作所 | Information recommendation device |
| WO2015055094A1 (en) * | 2013-10-15 | 2015-04-23 | 北京百度网讯科技有限公司 | Method and device for providing screening conditions and method and device for searching |
| CN105989004A (en) * | 2015-01-27 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information releasing pretreatment method and device |
| CN106228386A (en) * | 2016-07-12 | 2016-12-14 | 腾讯科技（深圳）有限公司 | A kind of information-pushing method and device |
| CN106296257A (en) * | 2015-06-11 | 2017-01-04 | 苏宁云商集团股份有限公司 | A kind of fixation of advertisement position put-on method based on user behavior analysis and system |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN107220852A (en) * | 2017-05-26 | 2017-09-29 | 北京小度信息科技有限公司 | Method, device and server for determining target recommended user |
| CN107463675A (en) * | 2017-08-03 | 2017-12-12 | 北京京东尚科信息技术有限公司 | Data processing method and its system |
| CN107644036A (en) * | 2016-07-21 | 2018-01-30 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and system of data object push |
-
2018
- 2018-02-02 CN CN201810105294.2A patent/CN110109901B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102208087A (en) * | 2010-03-30 | 2011-10-05 | 株式会社日立制作所 | Information recommendation device |
| WO2015055094A1 (en) * | 2013-10-15 | 2015-04-23 | 北京百度网讯科技有限公司 | Method and device for providing screening conditions and method and device for searching |
| CN105989004A (en) * | 2015-01-27 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information releasing pretreatment method and device |
| CN106296257A (en) * | 2015-06-11 | 2017-01-04 | 苏宁云商集团股份有限公司 | A kind of fixation of advertisement position put-on method based on user behavior analysis and system |
| CN106228386A (en) * | 2016-07-12 | 2016-12-14 | 腾讯科技（深圳）有限公司 | A kind of information-pushing method and device |
| CN107644036A (en) * | 2016-07-21 | 2018-01-30 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and system of data object push |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN107220852A (en) * | 2017-05-26 | 2017-09-29 | 北京小度信息科技有限公司 | Method, device and server for determining target recommended user |
| CN107463675A (en) * | 2017-08-03 | 2017-12-12 | 北京京东尚科信息技术有限公司 | Data processing method and its system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110109901A (en) | 2019-08-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110109901B (en) | Method and device for screening target object | |
| CN110135878B (en) | Method and device for determining sales price | |
| CN110020162B (en) | User identification method and device | |
| CN110135871B (en) | Method and device for calculating user repurchase period | |
| CN113763093A (en) | User portrait-based item recommendation method and device | |
| CN109961299A (en) | The method and apparatus of data analysis | |
| CN109840788B (en) | Method and device for analyzing user behavior data | |
| CN110473043B (en) | Article recommendation method and device based on user behaviors | |
| CN110738508B (en) | Data analysis method and device | |
| CN112784212B (en) | Inventory optimization method and device | |
| CN112749323B (en) | Method and device for constructing user portrait | |
| CN110866625A (en) | Promotion index information generation method and device | |
| US10467654B2 (en) | Forecasting customer channel choice using cross-channel loyalty | |
| CN108985805B (en) | Method and device for selectively executing push task | |
| CN110490682B (en) | Method and device for analyzing commodity attributes | |
| CN110796520A (en) | Commodity recommendation method and device, computing equipment and medium | |
| CN115936764A (en) | Product promotion method and device | |
| CN110838019A (en) | Method and device for determining trial supply distribution crowd | |
| CN110895777A (en) | Method and device for processing overtime service list | |
| CN115619452A (en) | User operation method and device based on arithmetic expression configuration | |
| CN110956514A (en) | Method and device for generating order information | |
| CN110766431A (en) | Method and device for judging whether user is sensitive to coupon | |
| CN110544140A (en) | method and device for processing browsing data | |
| CN113378043A (en) | User screening method and device | |
| Liu et al. | A genetic algorithm heuristic approach to general outsourcing capacitated production planning problems |
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 |