Search engine user satisfaction evaluation method based on mouse interaction sequence regional behavior joint modeling

Abstract

The invention discloses a search engine user satisfaction evaluation method based on mouse interaction sequence regional behavior joint modeling, which specifically comprises the following steps: 1) converting mouse cursor coordinates in a mouse interaction log into regional identifiers, and extracting a mouse interaction sequence by utilizing regions and behaviors; 2) introducing a data enhancement strategy based on multi-factor disturbance to perform data enhancement on the mouse interaction sequence; 3) introducing a regional behavior long-short-term memory network to learn feature representation of the mouse interaction sequence, and establishing a search engine user satisfaction classifier; and 4) identifying the user satisfaction label of the given sample by using the constructed user satisfaction classifier. The method for evaluating the user satisfaction of the search engine can promote the development of the search engine technology, so that higher-quality search service is provided for the user.

Description

technical field [0001] The invention relates to the field of Internet information technology, in particular to a search engine user satisfaction evaluation method based on mouse interaction sequence area behavior joint modeling. Background technique [0002] Search satisfaction is one of the important indicators to evaluate the quality of search engines based on user search experience. The evaluation results of user search satisfaction can bring the most intuitive performance description for the commercial operation of search engines, improve the ranking algorithm of search engines, improve Aspects such as user loyalty, increasing market share, etc. are crucial. [0003] Collecting explicit feedback on search satisfaction directly from users is too expensive and difficult to implement on a large scale, while mouse interaction logs record a large amount of interaction information between users and search engine results pages, such as users moving the mouse, rolling the mouse ...

AI Technical Summary

Problems solved by technology

However, in the mouse interaction sequence extracted by this method, the spatial information of the mouse cursor movement track in the mouse interaction log is captured by the anchor element. If the user's mouse cursor does not trigger any anchor element, the extracted mouse interaction sequence will not contain the spatial information of the mouse cursor movement trajectory, so some useful mouse cursor movement patterns will be lost

[0006] In addition, deep learning methods require a large amount of labeled data. Due to the expensive and time-consuming process of collecting labeled data, the trained model may be overfitted due to insufficient training data.

Images