Method for judging user emotion according to browsing history and expression recognition

Abstract

A method for judging the emotion of a user according to browsing history and expression recognition comprises the following steps: recording different indexes F of the emotion for a single page, wherein F can comprise n emotion states; obtaining and calculating an emotion set FP obtained by user browsing page residence time; obtaining the emotion value set FC of the number of times of repeatedly checking of browsing the page within the set time; obtaining an emotion value FE corresponding to the facial expression browsed in each single page (including repeatedly browsed); obtaining a result FWof the weather emotion state value in a specified range; obtaining results FT of the holiday emotion state values within a predetermined time range; and screening the statistical results of the n emotion states under various factors by using a decision tree method, and performing calculating to obtain the final probability distribution of the current emotion category of the user. The whole flow calculation is simple and clear and can be completed without the aid of special hardware equipment, the prediction effect is good, the requirement for data sources is simpler, and extremely high flexibility is achieved.

Description

technical field [0001] The invention relates to the technical field of user emotion estimation, in particular to a method for judging user emotion based on browsing history and facial expression recognition. Background technique [0002] In the era of intelligence, accurate browsing and pushing of information is emphasized, and different users often have different needs. In the traditional information content recommendation system, the information under this category or certain categories is often selectively recommended according to the degree of interest of different users in different types of information. If a user is interested in information of the "technology" category, certain information belonging to the category of "technology" is preferentially recommended when selecting information recommendation. This traditional approach has been widely used in past recommendation systems. By filtering recommendations based on different categories, users can better receive the...

AI Technical Summary

Problems solved by technology

In the method of purely using facial expressions for emotional inference, it is usually impossible to effectively judge the overall value of the user's emotion in a certain period of time

The use of special equipment for emotional inference can achieve higher accuracy, but due to the particularity of the equipment, there are high requirements for application scenarios and hardware equipment, and this kind of equipment that deeply analyzes changes in human muscle or brain waves is in the market. There are also certain security risks in the use of security

Images