Paper: Utilizing Variability of Time and Term Content within and across Users in Session Detection

ACL ID C10-2138
Title Utilizing Variability of Time and Term Content within and across Users in Session Detection
Venue International Conference on Computational Linguistics
Session Poster Session
Year 2010
Authors

In this paper, we describe a SVM classi- fication framework of session detection task on both Chinese and English query logs. With eight features on the aspects of temporal and content information ex- tracted from pairs of successive queries, the classification models achieve signifi- cantly superior performance than the stat- of-the-art method. Additionally, we find through ROC analysis that there exists great discrimination power variability among different features and within the same feature across different users. To fully utilize this variability, we build lo- cal models for individual users and com- bine their predictions with those from the global model. Experiments show that the local models do make significant im- provements to the global model, although the amount i...