CN113743972A - Method and device for generating article information - Google Patents
Method and device for generating article information Download PDFInfo
- Publication number
- CN113743972A CN113743972A CN202010826102.4A CN202010826102A CN113743972A CN 113743972 A CN113743972 A CN 113743972A CN 202010826102 A CN202010826102 A CN 202010826102A CN 113743972 A CN113743972 A CN 113743972A
- Authority
- CN
- China
- Prior art keywords
- article
- provider
- value
- promoted
- calculating
- 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 58
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 25
- 238000013145 classification model Methods 0.000 claims abstract description 22
- 230000002159 abnormal effect Effects 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 11
- 238000001514 detection method Methods 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 239000000126 substance Substances 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 238000006243 chemical reaction Methods 0.000 abstract description 14
- 230000000694 effects Effects 0.000 abstract description 13
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 10
- 238000004891 communication Methods 0.000 description 6
- 235000013312 flour Nutrition 0.000 description 5
- 230000003287 optical effect Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000000670 limiting effect Effects 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 230000001737 promoting effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002147 killing effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0242—Determining effectiveness of advertisements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0263—Targeted advertisements based upon Internet or website rating
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0277—Online advertisement
Abstract
The invention discloses a method and a device for generating article information, and relates to the technical field of computers. One embodiment of the method comprises: the basic value of the article can be calculated based on a back propagation algorithm model by acquiring multi-dimensional characteristic data of the article to be promoted; calculating the grade value of the article provider based on a classification model by acquiring the multi-dimensional characteristics of the article provider; further calculating the advertisement score of the article according to the basic score of the article, the grade score of the provider and the weight value of the article, generating article information to be promoted, and sending the article information; the client generates an article promotion page according to the article information; by combining the characteristic data of the articles, the article provider and the multi-dimensional dynamic characteristic data, the accuracy of calculating and selecting the promoted articles is improved, and the conversion effect of promotion is further improved; the scores of the articles and the providers are calculated through different models, the promotion score is obtained, and the accuracy of generating the information of the articles to be promoted is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a method and a device for generating article information.
Background
Currently, when an article provider promotes an article, the promotion effect is often evaluated according to the article sale condition, for example, in the e-commerce field, an advertisement form of charging according to an order amount is generally adopted; the method is also applied to the service industry, the advertisement form focuses on the advertisement conversion effect, in order to improve the advertisement conversion effect, the advertisement conversion effect of an article needs to be evaluated according to the related characteristics of the article, and then a suitable popularization article is selected, and in the existing evaluation advertisement conversion model, generally selected article characteristic indexes usually comprise static characteristics; the advertisement conversion model is generally obtained by training the characteristics of the articles by using a single model such as a logistic regression model or a neural network model. In the process of implementing the invention, the inventor finds that at least the following problems exist in the prior art:
in the existing advertisement conversion model, when the characteristics of the model are selected, the static values of the characteristics are usually selected to determine the promotion objects, and the accuracy of selecting the advertisement objects is low due to the fact that the dynamic characteristic values of the objects are not considered, and therefore the advertisement conversion effect is low. In addition, the existing advertisement conversion model generally adopts a single model for training, and when the characteristics of the articles are more, the single model is difficult to deal with the characteristics of the articles with multiple dimensions and complex scenes, so that the complexity of generating the information of the articles to be popularized is improved, and the calculation accuracy is reduced.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and an apparatus for generating article information, which are capable of calculating a basic score of an article based on a back propagation algorithm model by obtaining multi-dimensional feature data of the article to be promoted; calculating the grade value of the article provider based on a classification model by acquiring the multi-dimensional characteristics of the article provider; further calculating the advertisement score of the article according to the basic score of the article, the grade score of the provider and the weight value of the article, generating article information to be promoted, and sending the article information; the client generates an article promotion page according to the article information; by combining the characteristic data of the articles, the article provider and the multi-dimensional dynamic characteristic data, the accuracy of calculating and selecting the promoted articles is improved, and the conversion effect of promotion is further improved; the scores of the articles and the providers are calculated through different models, the promotion score is obtained, and the accuracy of generating the information of the articles to be promoted is improved.
To achieve the above object, according to an aspect of an embodiment of the present invention, there is provided a method of generating item information, including: acquiring the category of an article to be promoted, and calculating the basic score of the article to be promoted based on a back propagation algorithm model according to the characteristic data of the article contained in the category and the characteristic data of the article to be promoted; acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider; determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider; and forming a corresponding to-be-promoted article sequence based on the promotion score sequence, acquiring a set number of articles from the to-be-promoted article sequence to generate corresponding article information, and sending the article information.
Optionally, the method of generating item information,
the characteristic data of the article to be promoted comprises: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; and inputting the current characteristic value, the first dimension characteristic value and the second dimension characteristic value into a back propagation algorithm model to obtain a basic score of the article to be promoted.
Optionally, the method of generating item information,
obtaining the category of the article to be promoted, calculating the characteristic average value of the article contained in the category in a set historical time range, and calculating to obtain the first-dimension characteristic value based on the current characteristic value and the characteristic statistic value of the article to be promoted.
Optionally, the method of generating item information,
calculating a characteristic statistic value of the article to be promoted within a set time, and calculating to obtain the second dimension characteristic value based on the current characteristic value of the article to be promoted and the characteristic statistic value.
Optionally, the method of generating item information,
features of the item provider include: a provider characteristic value, a first dimension provider characteristic value and a second dimension provider characteristic value; and inputting the provider characteristic value, the first dimension provider characteristic value and the second dimension provider characteristic value into the classification model to obtain the grade value of the article provider.
Optionally, the method of generating item information,
the method comprises the steps of obtaining the category of the item provider, calculating a provider characteristic value statistic value of the item provider corresponding to the category within a set historical time range, and calculating to obtain the first dimension provider characteristic value based on the provider characteristic value of the item provider and the provider characteristic value statistic value.
Optionally, the method of generating item information,
and calculating the statistic value of the provider characteristic value of the item provider in a set time, and calculating to obtain the second dimension provider statistic value based on the provider characteristic value of the item provider and the statistic value of the provider characteristic value.
Optionally, the method of generating item information,
acquiring an abnormal characteristic value of the article to be promoted by using an abnormal value detection model, and inputting the abnormal characteristic value into a convolutional neural network model to acquire a weight value of the article to be promoted; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted.
To achieve the above object, according to a second aspect of an embodiment of the present invention, there is provided a method of generating item information, including: receiving the one or more item information, wherein the item information comprises a promotion score; and determining the display position of the corresponding article on the page according to the promotion score.
To achieve the above object, according to a third aspect of an embodiment of the present invention, there is provided an apparatus for article information, including: the system comprises an article score calculating module, a promotion score calculating module and an article information generating module; wherein the content of the first and second substances,
the item score calculating module is used for acquiring the category of the item to be promoted, and calculating the basic score of the item to be promoted based on a back propagation algorithm model according to the characteristic data of the item and the characteristic data of the item to be promoted contained in the category; acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider;
the promotion score calculation module is used for determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider;
and the article information generating module is used for forming a corresponding article sequence to be popularized based on the popularization score sequencing, acquiring a set number of articles from the article sequence to be popularized to generate corresponding article information, and sending the article information.
Optionally, the apparatus for generating item information is characterized in that,
the characteristic data of the article to be promoted comprises: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; and inputting the current characteristic value, the first dimension characteristic value and the second dimension characteristic value into a back propagation algorithm model to obtain a basic score of the article to be promoted.
Optionally, the apparatus for generating item information is characterized in that,
obtaining the category of the article to be promoted, calculating the characteristic average value of the article contained in the category in a set historical time range, and calculating to obtain the first-dimension characteristic value based on the current characteristic value and the characteristic statistic value of the article to be promoted.
Optionally, the apparatus for generating item information is characterized in that,
calculating a characteristic statistic value of the article to be promoted within a set time, and calculating to obtain the second dimension characteristic value based on the current characteristic value of the article to be promoted and the characteristic statistic value.
Optionally, the apparatus for generating item information is characterized in that,
features of the item provider include: a provider characteristic value, a first dimension provider characteristic value and a second dimension provider characteristic value; and inputting the provider characteristic value, the first dimension provider characteristic value and the second dimension provider characteristic value into the classification model to obtain the grade value of the article provider.
Optionally, the apparatus for generating item information is characterized in that,
the method comprises the steps of obtaining the category of the item provider, calculating a provider characteristic value statistic value of the item provider corresponding to the category within a set historical time range, and calculating to obtain the first dimension provider characteristic value based on the provider characteristic value of the item provider and the provider characteristic value statistic value.
Optionally, the apparatus for generating item information is characterized in that,
and calculating the statistic value of the provider characteristic value of the item provider in a set time, and calculating to obtain the second dimension provider statistic value based on the provider characteristic value of the item provider and the statistic value of the provider characteristic value.
Optionally, the apparatus for generating item information is characterized in that,
acquiring an abnormal characteristic value of the article to be promoted by using an abnormal value detection model, and inputting the abnormal characteristic value into a convolutional neural network model to acquire a weight value of the article to be promoted; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted.
In order to achieve the above object, according to a fourth aspect of an embodiment of the present invention, there is provided an apparatus for article information, comprising: a page generation module; the page generation module is used for receiving one or more item information, wherein the item information comprises a promotion score; and determining the display position of the corresponding article on the page according to the promotion score.
To achieve the above object, according to a fifth aspect of an embodiment of the present invention, there is provided an electronic device that generates item information, including: one or more processors; a storage device for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement a method as in any one of the methods of generating item information described above.
To achieve the above object, according to a sixth aspect of the embodiments of the present invention, there is provided a computer readable medium having a computer program stored thereon, characterized in that the program, when executed by a processor, implements the method as any one of the methods of generating article information described above.
One embodiment of the above invention has the following advantages or benefits: the basic value of the article can be calculated based on a back propagation algorithm model by acquiring multi-dimensional characteristic data of the article to be promoted; calculating the grade value of the article provider based on a classification model by acquiring the multi-dimensional characteristics of the article provider; further calculating the advertisement score of the article according to the basic score of the article, the grade score of the provider and the weight value of the article, generating article information to be promoted, and sending the article information; the client generates an article promotion page according to the article information; by combining the characteristic data of the articles, the article provider and the multi-dimensional dynamic characteristic data, the accuracy of calculating and selecting the promoted articles is improved, and the conversion effect of promotion is further improved; the scores of the articles and the providers are calculated through different models, the promotion score is obtained, and the accuracy of generating the information of the articles to be promoted is improved.
Further effects of the above-mentioned non-conventional alternatives will be 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 flow chart diagram of a method for generating item information according to an embodiment of the present invention;
FIG. 2 is a flow chart illustrating a method for obtaining a base score of an item according to an embodiment of the present invention;
FIG. 3 is a flow diagram illustrating a method for obtaining item provider tier scores according to one embodiment of the present invention;
fig. 4 is a schematic structural diagram of an apparatus for generating item information according to an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an apparatus for receiving information about an item according to an embodiment of the present invention;
FIG. 6 is an exemplary system architecture diagram in which embodiments of the present invention may be employed;
fig. 7 is a schematic block diagram of a computer system suitable for use in implementing a terminal device or server of an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present invention are described below with reference to the accompanying drawings, in which various details of embodiments of the invention are included to assist understanding, and which are to be considered as 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.
As shown in fig. 1, an embodiment of the present invention provides a method for generating item information, which may include the following steps:
step S101: acquiring the category of an article to be promoted, and calculating the basic score of the article to be promoted based on a back propagation algorithm model according to the characteristic data of the article contained in the category and the characteristic data of the article to be promoted; and acquiring the characteristics of an article provider of the article to be popularized, and calculating the grade value of the article provider based on a classification model according to the characteristics of the article and the characteristics of the article provider contained in the category.
Specifically, when determining an article to be promoted, first, a category of the article to be promoted is obtained, that is, a category of the article to be promoted is obtained, for example, the article to be promoted may be dishes of a restaurant, products of a travel agency, commodities of an electronic mall, and the like, and the category of the article is described below by taking the commodities of the electronic mall as an example, for example, a commodity category in the electronic mall such as a tool book in a book, flour in food, a television in a household appliance, a basketball in sports, a tent in the open air, and the like, and further, feature data of the article and feature data of the article to be promoted are included; wherein the characteristic data includes: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; the current characteristic value can be current sales information of an article to be promoted, taking a commodity as an example, and the current sales information of the article can be current price, recent sales volume, current commission, current preferential price and the like of the commodity; the first dimension characteristic value is calculated by the current information of the article to be promoted and the information of the articles of the same category, for example: calculating the average value of the prices of all the articles in the category in the past 30 days (namely, in a set historical time range), wherein the average value is an example of a characteristic statistic value of the articles in the set historical time range, and it can be understood that the statistic value can be an average value, a median, a percentile value and the like; further, dividing the current price (namely the current characteristic value) of the article to be promoted by the statistical value to obtain a first dimension characteristic value; it can be understood that the operation method for obtaining the first-dimension characteristic value by calculation may be division, deviation of statistical value, and the like, that is, obtaining the category of the article to be promoted, calculating a characteristic average value of the article included in the category within a set historical time range, and obtaining the first-dimension characteristic value by calculation based on the current characteristic value of the article to be promoted and the characteristic statistical value. The second dimension characteristic value is calculated by the current information of the article to be promoted and the historical information of the article to be promoted, for example: calculating the current price (namely the current characteristic value) of the article to be promoted divided by the average price (namely the characteristic statistic value) in the past 15 days (namely the set time) to be used as a second dimension characteristic value; it can be understood that the operation method for calculating the second dimension characteristic value may be division, obtaining statistical value deviation, and the like; namely, calculating a feature statistic value of the article to be promoted within a set time, and calculating the second dimension feature value based on the current feature value of the article to be promoted and the feature statistic value.
Further, calculating a basic score of the to-be-promoted item based on a back propagation algorithm model according to the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; the description of calculating the basic score of the to-be-promoted item by using the back propagation algorithm model is consistent with the steps S201 to S203, and is not repeated herein.
Further, acquiring characteristics of an article provider of the article to be promoted; features of the item provider include: a provider characteristic value, a first dimension provider characteristic value and a second dimension provider characteristic value; the provider feature value is service information of the provider, for example: provider good scores, provider after-sales scores, provider logistics scores, and the like; the first dimension provider feature value is calculated from the current information of the item provider and the information of all the item providers of the same category, for example: calculating the average value (namely the characteristic value of the provider) of the good scores of all the providers of the articles in the category within the last 30 days (namely the set historical time range); dividing the good scoring value (namely the characteristic value of the provider) of the article to be promoted by the statistical value to obtain a characteristic value of a first-dimension provider; it can be understood that the operation method for calculating the first dimension provider characteristic value may be division, obtaining statistical value deviation, and the like; that is, the category of the associated merchant is acquired, the provider feature value statistical value of the provider corresponding to the category within a set historical time range is calculated, and the first dimension provider feature value is calculated based on the provider feature value of the item provider and the provider feature value statistical value.
The second dimension provider feature value is calculated by the current information of the provider and the history information of the provider, for example: calculating the logistics score (namely the provider characteristic value) of the provider divided by the average logistics score (namely the provider characteristic value statistical value) in the past 7 days (namely the set time) as a second-dimension provider characteristic value; it can be understood that the operation method for calculating the second dimension provider characteristic value may be division, obtaining statistical value deviation, and the like; namely, a provider characteristic value statistical value of the merchant to be promoted within a set time is calculated, and the second-dimension provider characteristic value is calculated based on the provider characteristic value of the item provider and the provider characteristic value statistical value.
Further, the provider feature value, the first dimension provider feature value, and the second dimension provider feature value are input into the classification model, and the grade value of the item provider is obtained. The description of calculating the rating values of the providers based on the classification model is consistent with steps S301 to S303, and is not repeated here.
It can be understood that the specific contents of the features and feature values of the to-be-promoted item, the specific calculation methods of the first dimension feature value and the second dimension feature value, the specific contents of the features and feature values of the item provider, the specific calculation methods of the first dimension provider feature value and the second dimension provider feature value, the specific calculation methods of the set history time range, the set time and the statistical value are set according to the industry, the service scene and the service requirement of the item provider, and the present invention does not limit the contents.
Step S102: and determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider.
Specifically, according to the description of step S101, the base score of the item to be promoted and the grade value of the item provider are obtained, and further, the promotion score of the item to be promoted is determined based on the base score of the item to be promoted and the grade value of the item provider.
For example: the promotion score is calculated using equation (1) as an example, as follows:
ComprehensiveScore＝ShopScore*BaseScore(1)
wherein the basecore is indicated as a base score of the item to be promoted; ShopScore indicates the rating value for the provider of the item; "x" indicates a multiplication symbol, and comprehensionscore indicates an item provider score for the item to be promoted; namely, calculating a basic score of an article to be promoted and a grade value of an article provider, and determining a promotion score of the article to be promoted; the invention does not limit the concrete formula for calculating the advertisement score of the to-be-promoted item based on the basic score of the to-be-promoted item and the grade value of the item provider.
Further, another embodiment of determining a promotional score is: acquiring an abnormal characteristic value of the article to be promoted by using an abnormal value detection model, and inputting the abnormal characteristic value into a convolutional neural network model to acquire a weight value of the article to be promoted; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted.
Specifically, for example: in the promotion activity of item second killing or promotion strength, etc., some items may have abnormal characteristic values, for example, the commission is greatly increased or the price is greatly reduced, and the advertisement of the part of items can generally obtain higher conversion, therefore, the commodity with commission increase or price reduction can be detected in advance by using the abnormal characteristic value detection model, namely, the abnormal characteristic values such as commission change, price change or benefit strength change can be identified in advance (for example, one day in advance) by using the abnormal value detection model based on the static characteristic and dynamic characteristic of commission and price, further, the corresponding items can be obtained, the weight values of the part of items can be determined by using the convolutional neural network model, and the weight values are set to be more than 1, and the weight values of the items without abnormal value detection are set to be 1. By the steps, the characteristic value of the article can be dynamically acquired, and the promotion score of the article to be promoted can be accurately acquired.
Further, calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted, taking the following formula as an example:
ComprehensiveScore＝ShopScore*BaseScore*WeightScore
wherein the basecore is indicated as a base score of the item to be promoted; ShopScore indicates the rating value for the provider of the item; the WeightScore indicates the weight value of the item to be promoted; "" indicates a multiplication symbol, Compressenstive score indicates an advertisement score for the item to be promoted; acquiring an abnormal characteristic value of the article to be popularized by using an abnormal value detection model, and inputting the abnormal characteristic value into a convolutional neural network model to acquire a weight value of the article to be popularized; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted. The specific formula for calculating the promotion score of the to-be-promoted item based on the basic score, the level value and the weight value of the associated item provider of the to-be-promoted item is not limited.
Step S103: and forming a corresponding to-be-promoted article sequence based on the promotion score sequence, acquiring a set number of articles from the to-be-promoted article sequence to generate corresponding article information, and sending the article information.
Specifically, according to the description of step S102, the promotion scores of a plurality of items to be promoted are obtained through calculation; further, ranking the promotional scores, e.g., from high to low; and acquiring the articles corresponding to the set number of promotion scores from high to first as promotion commodities. It can be understood that one or more sorted items to be promoted may belong to the same item provider, or may belong to different item providers (e.g., a certain business alliance advertises, or a certain company advertises multiple businesses under the banner); further, a set number of articles are obtained from the article sequence to be promoted to generate corresponding article information, and the article information is sent to a client. It will be appreciated that the set number is determined by the ad design, for example: the set number can be 1, the set number can be 50, and the set number of the displayed advertisement commodities is related to the contents, the form, the duration, the display mode and other factors of the displayed and interface advertisements; further, the client determines the display position of the article on the page and displays the article according to the promotion score contained in the received article information, namely, the article information is received, and the article information contains the promotion score; and determining the display position of the corresponding article on the page according to the promotion score. The invention does not limit the specific numerical values of the set quantity, the specific contents corresponding to the promoted articles, the display mode and the specific contents of the interface.
As shown in FIG. 2, an embodiment of the present invention provides a method for obtaining a base score of an advertised item, which may include the following steps:
step S201: acquiring the category of an article to be promoted, and calculating the basic score of the article to be promoted based on a back propagation algorithm model according to the characteristic data of the article contained in the category and the characteristic data of the article to be promoted; the characteristic data of the commodity to be promoted comprises: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value.
Specifically, regarding obtaining the category of the article to be promoted, the description of the article-containing feature data according to the category and the feature data of the article to be promoted is consistent with step S101, and is not repeated here; further, calculating the candidate vector based on a back propagation algorithm modelThe following illustrates the steps of calculating the basic score of the item to be promoted based on a back propagation algorithm model: specifically, the feature data of the article to be promoted includes: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value are obtained, for example, commissions proportions, prices and preferential prices of the items to be promoted are obtained as the current characteristic values, and the first dimension characteristic value and the second dimension characteristic value are calculated based on the current characteristic values; and then training the characteristic value of the article to be popularized by adopting a back propagation algorithm model, and obtaining the basic score of the article to be popularized. It can be understood that commission, commission proportion, price and preferential price are the basic characteristics for determining the advertisement conversion effect, so that the characteristic information engineering construction of the article is carried out according to the commission, commission proportion, price and preferential price, the back propagation algorithm model is adopted for training, furthermore, in order to prevent the back propagation algorithm model from generating the overfitting phenomenon, the 'regularization' strategy can be adopted, and the square sum of the connection right and the threshold value can be added into the error objective function
This term describes the complexity of the network, thereby avoiding over-fitting due to the back-propagation algorithm model being too complex. The error objective function of the back propagation algorithm model after being processed by the regularization strategy is shown in formula (2).
Wherein
Indicates that the network is in (x)k,yk) The error in (2); w is aiRepresents the connection weight and the threshold; λ and 1- λ represent the weights of the empirical error and the complexity of the network, respectively. Wherein, λ ∈ (0,1) is used to trade off the two terms of the empirical error and the network complexity, and preferably, the effect is better when λ is set to 0.7 according to the test.
Step S202: obtaining the category of the article to be promoted, calculating the characteristic statistic value of the article contained in the category in a set historical time range, and calculating to obtain the first-dimension characteristic value based on the current characteristic value and the characteristic statistic value of the article to be promoted.
Specifically, the method of calculating the first-dimension feature value is described below in the example shown in table 1: taking the price of the article to be promoted as an example, obtaining the current characteristic value (for example, the current price) of the article to be promoted, further calculating the characteristic statistic value of the article contained in the corresponding category in the set historical time range, for example, calculating the average price (namely, the characteristic statistic value) of all the articles in the category within the past 30 days (namely, in the set historical time range), and calculating the first-dimension characteristic value by dividing the current price by the average price of all the articles; for example: the current price of the flour A is 110 yuan, the price of the flour of the same category (including the same weight) in the last 30 days is calculated to be 100 yuan, and the first dimension characteristic value of the price is obtained by dividing the calculated price by 110 by 100; as shown in table 1, similarly to the price, according to the current feature value: and calculating corresponding first dimension characteristic values of commissions, commissions ratios, preferential prices and sales volumes.
Table 1 example of calculating characteristic values of an item to be promoted
Step S203: calculating a characteristic statistic value of the article to be promoted within a set time, and calculating to obtain the second dimension characteristic value based on the current characteristic value of the article to be promoted and the characteristic statistic value.
Specifically, the method of calculating the second-dimension feature value is described still by the example shown in table 1: taking the price of the article to be promoted as an example, obtaining the current characteristic value (for example, the current price) of the article to be promoted, further calculating a characteristic statistic value of the article to be promoted within a set time, for example, calculating an average price (as an example of the characteristic statistic value) of the article to be promoted within the last 7 days (as the set time), further dividing the average price (as the characteristic statistic value) by the current price, and calculating a second dimension characteristic value; for example: the current price of the flour A is 110 yuan, the price of the flour A in the last 7 days is calculated to be 120 yuan, and the second dimension characteristic value corresponding to the price is obtained by dividing the calculated price by 120 through 110; as shown in table 1, similarly to the price, according to the current feature value: and calculating corresponding second dimension characteristic values of commissions, commission ratio, preferential price and sales volume.
As shown in fig. 3, an embodiment of the present invention provides a method for obtaining a tier score of an item provider, which may include the following steps:
step S301: acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the classification including the characteristics of the article provider; features of the item provider include: features of the item provider include: a provider feature value, a first dimension provider feature value, and a second dimension provider feature value.
Specifically, regarding the category of the acquired to-be-promoted item, the description of the acquired category including the characteristics of the item provider is consistent with that in step S101, and is not repeated here; further, the grade value of the item provider is calculated based on the classification model, and the following illustrates the steps of calculating the grade value of the item provider based on the classification model, for example: using the softmax model as the classification model, the cost function formula corresponding to the softmax model is shown as formula (3):
wherein m indicates the number of samples; k indicates the classification number (for example, the grade value of the article provider is set to five grade values of upper grade, middle grade, lower grade and the like, so that k is 5); 1 {. is indicated as an illustrative function; y is indicated as a label, in this example as an item provider rating value; x indicates a provider feature value, a first dimension provider feature value, a second dimension provider feature value, for example, 9 feature values shown in table 2 contained in step S302 are taken as an example; θ is indicated as softmax model parameter vector; λ is denoted as the weight attenuation coefficient; the description of calculating the corresponding first dimension provider characteristic value and second dimension provider characteristic value based on the provider characteristic values (such as the good score value, the after-sales score value and the logistics score) is consistent with the steps S302-S303, and is not repeated herein; further, in the present embodiment, the levels of the provider are set to upper-upper, middle, lower, and lower (5 levels in total), that is, k is 5, and the level values corresponding to the levels are set to 1.4, 1.2, 1.0, 0.8, and 0.6, respectively.
Alternatively, the above softmax model formula can be solved by using a gradient descent method, which needs to solve the derivative of the cost function, which can be shown as follows:
after a specific softmax model is obtained through solving, the probability of the grade of the provider can be predicted, and the grade score of the provider is further determined according to an example formula shown as the following formula:
the classification model used and the particular formula used to determine the provider rating score are not limiting of the present invention.
Step S302: the method comprises the steps of obtaining the characteristics of an article provider, calculating a provider characteristic value statistical value of a merchant corresponding to the category within a set historical time range, and calculating to obtain a first dimension provider characteristic value based on the provider characteristic value and the provider characteristic value statistical value of the associated provider.
Specifically, the method of calculating the first dimension provider feature value is described below in the example shown in table 2: taking the good scoring value of the provider as an example, obtaining a provider characteristic value (such as a good scoring value) of an article to be promoted, further calculating a characteristic statistic value of the provider in a set historical time range, which is contained in the corresponding category, for example, calculating an average value (namely the provider characteristic statistic value) of the good scoring values of all the providers in the category within the last 30 days (namely the set historical time range), and dividing the average value by the good scoring value to calculate a first dimension provider characteristic value corresponding to the good scoring value of the provider; for example: the good score value of the provider A is 4.9, the average value (namely the feature statistical value) of the good score values of the same class providers in the last 30 days is calculated to be 4, and the value is calculated to be 4 divided by 4 to obtain 1.23, namely the first dimension provider feature value of the good score values; as shown in table 2, similarly to the good score value, according to the provider feature value: and calculating corresponding first dimension characteristic values according to the after-sale value and the logistics value. That is, the first-dimension provider feature value is calculated based on the provider feature value of the item provider and the provider feature value statistical value.
Table 2 example of feature values for computing item providers
Step S303: and calculating a provider characteristic value statistical value of the item provider within a set time, and calculating the second dimension provider characteristic value based on the provider characteristic value and the provider characteristic value statistical value of the item provider.
Specifically, the method of calculating the second dimension provider feature value is still described in the example shown in table 2: taking the good scoring value of an article provider as an example, obtaining the article provider (for example, the good scoring value) of the article to be promoted, further calculating the feature value statistical value of the provider in a set time of the provider contained in the corresponding category, for example, calculating the average value (namely, the feature value statistical value of the provider) of the good scoring value of the provider in the last 7 days (namely, in the set time), and calculating the feature value of the second dimension provider corresponding to the good scoring value of the provider by dividing the good scoring value by the average value; for example: the good score value of the provider A is 4.9, the average good score value in the past 7 days is calculated to be 4, and the value is divided by 4 to be calculated to be 1.23, namely the characteristic value of the second-dimension provider with the good score value; as shown in table 2, similarly to the good score value, according to the provider feature value: and calculating corresponding second dimension provider characteristic values according to the after-sale scores and the logistics scores. That is, the second-dimension provider feature value is calculated based on the provider feature value of the item provider and the provider feature value statistical value.
As shown in fig. 4, an embodiment of the present invention provides an apparatus 400 for generating item information, including: a module 401 for calculating the score of an article, a module 402 for calculating the promotion score and a module 403 for generating information of the article; wherein the content of the first and second substances,
the item score calculating module 401 is configured to obtain a category of an item to be promoted, and calculate a basic score of the item to be promoted based on a back propagation algorithm model according to the category including characteristic data of the item and characteristic data of the item to be promoted; acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider;
the promotion score calculation module 402 determines a promotion score of the to-be-promoted item according to the basic score of the to-be-promoted item and the grade value of the item provider;
the item information generating module 403 forms a corresponding to-be-promoted item sequence based on the promotion score, obtains a set number of items from the to-be-promoted item sequence to generate corresponding item information, and sends the item information.
Optionally, the module 401 for calculating a score of an item, which includes feature data of the item to be promoted, includes: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; and inputting the current characteristic value, the first dimension characteristic value and the second dimension characteristic value into a back propagation algorithm model to obtain a basic score of the article to be promoted.
Optionally, the item score calculating module 401 is configured to obtain a category of the item to be promoted, calculate a feature statistic of the item included in the category within a set historical time range, and calculate the first-dimension feature value based on the current feature value of the item to be promoted and the feature statistic.
Optionally, the item score calculating module 401 is configured to calculate a feature statistic of the item to be promoted within a set time, and calculate the second dimension feature value based on the current feature value of the item to be promoted and the feature statistic.
Optionally, the module 401 for calculating an item score includes features of the item provider including: a provider characteristic value, a first dimension provider characteristic value and a second dimension provider characteristic value; and inputting the provider characteristic value, the first dimension provider characteristic value and the second dimension provider characteristic value into the classification model to obtain the grade value of the article provider.
Optionally, the module 401 for calculating an item score is configured to obtain a category of the item provider, calculate a provider feature value statistical value of a provider corresponding to the category within a set historical time range, and calculate the first dimension provider feature value based on the provider feature value of the item provider and the provider feature value statistical value.
Optionally, the module 401 for calculating an item score is configured to calculate a statistical value of a provider feature value of the item provider within a set time, and calculate the second dimension provider statistical value based on the provider feature value of the item provider and the statistical value of the provider feature value.
Optionally, the promotion score calculation module 402 is configured to obtain an abnormal feature value of the to-be-promoted item by using an abnormal value detection model, input the abnormal feature value into a convolutional neural network model, and obtain a weight value of the to-be-promoted item; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted.
As shown in fig. 5, an embodiment of the present invention provides an apparatus 500 for generating item information, including: a generate page module 501; the page generating module 501 is configured to receive one or more item information, where the item information includes a promotion score; and determining the display position of the corresponding article on the page according to the promotion score.
An embodiment of the present invention further provides an electronic device for generating article information, including: one or more processors; 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 provided by any one of the above embodiments.
Embodiments of the present invention further provide a computer-readable medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method provided in any of the above embodiments.
Fig. 6 illustrates an exemplary system architecture 600 of a method of item information or an apparatus of item information to which embodiments of the invention may be applied.
As shown in fig. 6, the system architecture 600 may include terminal devices 601, 602, 603, a network 604, and a server 605. The network 604 serves to provide a medium for communication links between the terminal devices 601, 602, 603 and the server 605. Network 604 may include various types of connections, such as wire, wireless communication links, or fiber optic cables, to name a few.
A user may use the terminal devices 601, 602, 603 to interact with the server 605 via the network 604 to receive or send messages or the like. Various client applications, such as a web browser application, a search-type application, an instant messaging tool, a mailbox client, and the like, may be installed on the terminal devices 601, 602, 603.
The terminal devices 601, 602, 603 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 605 may be a server providing various services, for example, the server transmits the generated item information to the terminal devices 601, 602, and 603, and the terminal devices determine the display positions of the items according to the received item information and generate pages.
The method for generating the article information according to the embodiment of the present invention is generally executed by the server 605, and accordingly, the apparatus for generating the article information is generally provided in the server 605.
It should be understood that the number of terminal devices, networks, and servers in fig. 6 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
Referring now to FIG. 7, shown is a block diagram of a computer system 700 suitable for use with a terminal device implementing an embodiment of the present invention. The terminal device shown in fig. 7 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 7, the computer system 700 includes a Central Processing Unit (CPU)701, which can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)702 or a program loaded from a storage section 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data necessary for the operation of the system 700 are also stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
The following components are connected to the I/O interface 705: an input portion 706 including a keyboard, a mouse, and the like; an output section 707 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 708 including a hard disk and the like; and a communication section 709 including a network interface card such as a LAN card, a modem, or the like. The communication section 709 performs communication processing via a network such as the internet. A drive 710 is also connected to the I/O interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that a computer program read out therefrom is mounted into the storage section 708 as necessary.
In particular, according to the embodiments of the present disclosure, the processes described above with reference to the 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 illustrated in the flow chart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 709, and/or installed from the removable medium 711. The computer program performs the above-described functions defined in the system of the present invention when executed by the Central Processing Unit (CPU) 701.
It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination 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 present invention, 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, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. 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 flowchart 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 and/or units described in the embodiments of the present invention may be implemented by software, and may also be implemented by hardware. The described modules and/or units may also be provided in a processor, and may be described as: a processor includes an item score calculation module, a promotion score calculation module, and a page information generation module. The names of the modules do not limit the modules, for example, the module for calculating the promotion score can be further described as a module for determining the promotion score of the item to be promoted according to the basic score of the item to be promoted and the grade score of the item provider.
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 separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by a device, cause the device to comprise: acquiring the category of an article to be promoted, and calculating the basic score of the article to be promoted based on a back propagation algorithm model according to the characteristic data of the article contained in the category and the characteristic data of the article to be promoted; acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider; determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider; and forming a corresponding to-be-promoted article sequence based on the promotion score sequence, acquiring a set number of articles from the to-be-promoted article sequence to generate corresponding article information, and sending the article information. Receiving one or more item information, wherein the item information comprises a promotion score; and determining the display position of the corresponding article on the page according to the promotion score.
According to the technical scheme of the embodiment of the invention, the basic score of the article can be calculated based on the back propagation algorithm model by acquiring the multi-dimensional characteristic data of the article to be promoted; calculating the grade value of the article provider based on a classification model by acquiring the multi-dimensional characteristics of the article provider; further calculating the advertisement score of the article according to the basic score of the article, the grade score of the provider and the weight value of the article, generating article information to be promoted, and sending the article information; the client generates an article promotion page according to the article information; by combining the characteristic data of the articles, the article provider and the multi-dimensional dynamic characteristic data, the accuracy of calculating and selecting the promoted articles is improved, and the conversion effect of promotion is further improved; the scores of the articles and the providers are calculated through different models, the promotion score is obtained, and the accuracy of generating the information of the articles to be promoted is improved.
The above-described embodiments should not be construed as limiting the scope of the invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions can occur, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (13)
1. A method of generating item information, comprising:
acquiring the category of an article to be promoted, and calculating the basic score of the article to be promoted based on a back propagation algorithm model according to the characteristic data of the article contained in the category and the characteristic data of the article to be promoted;
acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider;
determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider;
and forming a corresponding to-be-promoted article sequence based on the promotion score sequence, acquiring a set number of articles from the to-be-promoted article sequence to generate corresponding article information, and sending the article information.
2. The method of claim 1,
the characteristic data of the article to be promoted comprises: the current characteristic value, the first dimension characteristic value and the second dimension characteristic value; and inputting the current characteristic value, the first dimension characteristic value and the second dimension characteristic value into a back propagation algorithm model to obtain a basic score of the article to be promoted.
3. The method of claim 2,
obtaining the category of the article to be promoted, calculating the characteristic average value of the article contained in the category in a set historical time range, and calculating to obtain the first-dimension characteristic value based on the current characteristic value and the characteristic statistic value of the article to be promoted.
4. The method of claim 2,
calculating a characteristic statistic value of the article to be promoted within a set time, and calculating to obtain the second dimension characteristic value based on the current characteristic value of the article to be promoted and the characteristic statistic value.
5. The method of claim 1,
features of the item provider include: a provider characteristic value, a first dimension provider characteristic value and a second dimension provider characteristic value; and inputting the provider characteristic value, the first dimension provider characteristic value and the second dimension provider characteristic value into the classification model to obtain the grade value of the article provider.
6. The method of claim 5,
the method comprises the steps of obtaining the category of the item provider, calculating a provider characteristic value statistic value of the item provider corresponding to the category within a set historical time range, and calculating to obtain the first dimension provider characteristic value based on the provider characteristic value of the item provider and the provider characteristic value statistic value.
7. The method of claim 5,
and calculating the statistic value of the provider characteristic value of the item provider in a set time, and calculating to obtain the second dimension provider statistic value based on the provider characteristic value of the item provider and the statistic value of the provider characteristic value.
8. The method of claim 1,
acquiring an abnormal characteristic value of the article to be promoted by using an abnormal value detection model, and inputting the abnormal characteristic value into a convolutional neural network model to acquire a weight value of the article to be promoted; and calculating the promotion score of the article to be promoted according to the basic score of the article to be promoted, the grade value of the article provider and the weight value of the article to be promoted.
9. A method of generating item information, comprising:
receiving one or more item information, wherein the item information comprises a promotion score; and determining the display position of the corresponding article on the page according to the promotion score.
10. An apparatus for generating item information, comprising: the system comprises an article score calculating module, a promotion score calculating module and an article information generating module; wherein the content of the first and second substances,
the item score calculating module is used for acquiring the category of the item to be promoted, and calculating the basic score of the item to be promoted based on a back propagation algorithm model according to the characteristic data of the item and the characteristic data of the item to be promoted contained in the category; acquiring the characteristics of an article provider of the article to be promoted, and calculating the grade value of the article provider based on a classification model according to the category including the characteristics of the article and the characteristics of the article provider;
the promotion score calculation module is used for determining the promotion score of the article to be promoted according to the basic score of the article to be promoted and the grade value of the article provider;
and the article information generating module is used for forming a corresponding article sequence to be popularized based on the popularization score sequencing, acquiring a set number of articles from the article sequence to be popularized to generate corresponding article information, and sending the article information.
11. An apparatus for generating item information, comprising: a page generation module; wherein the content of the first and second substances,
the page generation module is used for receiving one or more item information, and the item information comprises promotion scores; and determining the display position of the corresponding article on the page according to the promotion score.
12. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-8 or claim 9.
13. A computer-readable medium, on which a computer program is stored, which program, when being executed by a processor, is adapted to carry out the method of any one of claims 1-8 or 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010826102.4A CN113743972A (en) | 2020-08-17 | 2020-08-17 | Method and device for generating article information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010826102.4A CN113743972A (en) | 2020-08-17 | 2020-08-17 | Method and device for generating article information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113743972A true CN113743972A (en) | 2021-12-03 |
Family
ID=78728023
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010826102.4A Pending CN113743972A (en) | 2020-08-17 | 2020-08-17 | Method and device for generating article information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113743972A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107103487A (en) * | 2017-03-02 | 2017-08-29 | 浙江兰德纵横网络技术股份有限公司 | A kind of advertisement sending method based on big data |
| CN107256494A (en) * | 2017-05-18 | 2017-10-17 | 深圳大学 | A kind of item recommendation method and device |
| CN108876422A (en) * | 2017-05-09 | 2018-11-23 | 北京京东尚科信息技术有限公司 | For the method, apparatus of information popularization, electronic equipment and computer-readable medium |
| CN109167816A (en) * | 2018-08-03 | 2019-01-08 | 广州虎牙信息科技有限公司 | Information-pushing method, device, equipment and storage medium |
| CN110704727A (en) * | 2019-08-30 | 2020-01-17 | 中国平安人寿保险股份有限公司 | Information pushing method and device and computer equipment |
| US20200074509A1 (en) * | 2018-08-31 | 2020-03-05 | Baidu Online Network Technology (Beijing) Co., Ltd. | Business data promotion method, device, terminal and computer-readable storage medium |
| CN111192108A (en) * | 2019-12-16 | 2020-05-22 | 北京淇瑀信息科技有限公司 | Sorting method and device for product recommendation and electronic equipment |
| CN111325587A (en) * | 2018-12-13 | 2020-06-23 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating information |
-
2020
- 2020-08-17 CN CN202010826102.4A patent/CN113743972A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107103487A (en) * | 2017-03-02 | 2017-08-29 | 浙江兰德纵横网络技术股份有限公司 | A kind of advertisement sending method based on big data |
| CN108876422A (en) * | 2017-05-09 | 2018-11-23 | 北京京东尚科信息技术有限公司 | For the method, apparatus of information popularization, electronic equipment and computer-readable medium |
| CN107256494A (en) * | 2017-05-18 | 2017-10-17 | 深圳大学 | A kind of item recommendation method and device |
| CN109167816A (en) * | 2018-08-03 | 2019-01-08 | 广州虎牙信息科技有限公司 | Information-pushing method, device, equipment and storage medium |
| US20200074509A1 (en) * | 2018-08-31 | 2020-03-05 | Baidu Online Network Technology (Beijing) Co., Ltd. | Business data promotion method, device, terminal and computer-readable storage medium |
| CN111325587A (en) * | 2018-12-13 | 2020-06-23 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating information |
| CN110704727A (en) * | 2019-08-30 | 2020-01-17 | 中国平安人寿保险股份有限公司 | Information pushing method and device and computer equipment |
| CN111192108A (en) * | 2019-12-16 | 2020-05-22 | 北京淇瑀信息科技有限公司 | Sorting method and device for product recommendation and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 任燕;王彪;: "信息推送方法及信息推送系统企业知识产权分析", 电脑知识与技术, no. 02, pages 281 - 283 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110363604B (en) | Page generation method and device | |
| US9911130B1 (en) | Attribution modeling using regression analysis | |
| CN109711917A (en) | Information-pushing method and device | |
| CN112446764A (en) | Game commodity recommendation method and device and electronic equipment | |
| CN113763019A (en) | User information management method and device | |
| CN114493786A (en) | Information recommendation method and device | |
| CN110827104A (en) | Method and device for recommending commodities to user | |
| CN110633405A (en) | Method and device for pushing information | |
| CN115293291B (en) | Training method and device for sequencing model, sequencing method and device, electronic equipment and medium | |
| CN112308477A (en) | Inventory positioning method and device | |
| CN110490682B (en) | Method and device for analyzing commodity attributes | |
| CN107357847B (en) | Data processing method and device | |
| CN115099865A (en) | Data processing method and device | |
| CN113743972A (en) | Method and device for generating article information | |
| CN115271866A (en) | Product recommendation method and device, electronic equipment and readable storage medium | |
| US20240078585A1 (en) | Method and apparatus for sharing information | |
| CN113778979A (en) | Method and device for determining live broadcast click rate | |
| CN113158032A (en) | Information pushing method and device | |
| CN109583922B (en) | Method and device for analyzing purchase demand | |
| CN112784861A (en) | Similarity determination method and device, electronic equipment and storage medium | |
| CN112418898A (en) | Article demand data analysis method and device based on multi-time window fusion | |
| CN111488537A (en) | Page display method and device | |
| CN113159877A (en) | Data processing method, device, system and computer readable storage medium | |
| CN111738789A (en) | Article information pushing method, device, equipment and storage medium | |
| CN107730320B (en) | User mining method and device, electronic equipment and computer readable storage medium |
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 |