Interest mining method for user browsing behaviors

Abstract

The invention discloses an interest mining method for user browsing behaviors. For users u1,u2 and u3, within an appointed time, the user u1 accesses tags t1, t2 and t3, the user u2 accesses tags t2, and the user t3 accesses tags t2 and t3. The method comprises the following steps: (1), certain typical websites inside each interest label are labeled, and the default of weight of interest corresponding to the labeled websites tag->interest is 1.0; (2), according two graph models established between users and website tags accessed by users within an appointed time and through n turns of random walk, wherein n refers to positive integers, the results of n turns of random walk are collected, and the weight of user-> tag can be calculated; (3), user->interest is obtained through the product of the user->tag obtained in the step (2) and the user->interest obtained in the step (3), and user->interest refers to a confidence coefficient of each user to the interest tag; (4), a threshold value a is arranged, and when the confidence coefficient of the user->interest is larger than a, the interest tag for the user is predicted.

Description

technical field [0001] The invention belongs to the technical field of big data processing and analysis, and in particular relates to an interest mining method of user browsing behavior. Background technique [0002] After the Internet has gradually entered the era of big data, with the in-depth research and application of big data technology, enterprises are increasingly focusing on using big data to portray "user portraits", and then dig deeper into potential business value. User interest mining can Dig out different interest groups to facilitate precise marketing services. [0003] Users browse a lot of websites when surfing the Internet, and digging out the user's interests and preferences from these numerous websites is interest mining. In the existing technology, method A is to mark some websites with interest, such as JD.com and Tmall websites corresponding to "e-commerce iQiyi and Youku Tudou correspond to the "audio-visual entertainment preference", set a threshold...

AI Technical Summary

Problems solved by technology

[0005] Method A is simple, effective, and easy to implement, but the effect is limited by the scale of the website labeling. For example, if only Jingdong and Tmall websites are marked as "e-commerce shopping preferences", then some users have visited "Vipshop", Websites such as "Gome" and "Suning" will be ignored, and will not correspond to "e-commerce shopping preferences", and it is impossible to label all e-commerce shopping websites manually. The larger the labeling scale, the higher the labor cost

[0006] Method B uses a supervised machine learning model to solve the problem of interest mining. First, a large number of user interest labeling samples are required. This is not easy to solve in most scenarios. Faced with the problem of cold start, the initial batch of user interest labeling Samples are not easy to obtain. In addition, when the information of the group of users who visit the website is updated over time, and when the user's interest changes, there will be problems in predicting the user's interest label based on the similarity between users.

Images