Individualized recommending method in combination of rating data and label data
A technology of labeling data and recommendation methods, which is applied in electronic digital data processing, special data processing applications, instruments, etc., and can solve problems such as affecting recommendation results, inaccurate similarity between users, and inaccurate prediction scores.
Inactive Publication Date: 2013-09-11
NANJING UNIV
View PDF2 Cites 1 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0011] The technical problem to be solved by the present invention is: in the existing user-based collaborative filtering method, when there are few products that are jointly rated by two users, the calculated similarity between users may be inaccurate, resulting in inaccurate predicted ratings , affecting the recommendation result
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0034] The features of the present invention are as follows:
[0035] 1) Clean the label data in advance to improve the quality of the label data;
[0036] 2) Simultaneously use the user's rating data on the product and the product's tagged label data, and effectively combine and calculate the user's rating data on the label;
[0037] 3) When calculating the user's rating data on tags in step (2), the relationship between tags is considered. The relationship between tags is expressed by calculating the probability distribution method of the co-occurrence of tags;
[0038] 4) Using the idea of user-based collaborative filtering to calculate the similarity between users according to the rating data of users on tags, and generate similar users of the target user.
[0039] The present invention combines the user's rating data on the product and the product's tag data to calculate and generate the user's rating data on the tag, and then uses the user-based collaborative filteri...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an individualized recommending method in combination of rating data and label data. The method comprises the following steps: generating the rating data of user to a label by computing the rating data of user to a product and the marked label data of the product; then computing the similarity among users by using a user-based collaborative filtering algorithm according to the rating data of user to the label so as to generate a similar user group of a target user; and finally predicating an unknown rating of the target user to the product according to the rating of the similar user group. By computing the similarity among the users in combination of the rating data and the label data, the computed similarity is more accurate, and finally the more accurate prediction rating for the target user is generated, and the recommending effect is improved.
Description
technical field [0001] The present invention belongs to the field of personalized recommendation, and calculates and generates user's rating data on tags by combining user's rating data on products and product tagged label data, and adopts user-based collaborative filtering idea to calculate user rating data based on user's rating data on tags. The similarity between them is used to find similar user groups of target users. The present invention is mainly applied to WEB application scenarios that have both user rating data on products and product tag data, and is a personalized recommendation method that combines rating data and tag data. Background technique [0002] With the continuous development of the Internet, personalized recommendation technology is becoming more and more important. Personalized recommendation technology can help users quickly find what they really need among massive products. User-based collaborative filtering is a very successful and widely used ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F17/30
Inventor 刘嘉祁奇陈振宇吴清王维清
Owner NANJING UNIV