Hierarchical Query Task Clustering using Search Logs

Computer Engineering

In this paper we have enhanced Query Task Clustering algorithm for to give additional feature such as URL recommendation and next sub-task based suggestion on top of existing query suggestion model. Hierarchical Query Task Clustering algorithms breaks user search task into sub tasks and captures relation between sub-tasks. It uses click through bipartite graph to generate query similarity without using any explicit judgements. It also uses query features such as average dwell time for query, total no of URLs clicks and no of unique URLs clicked to give more accurate results. Experimental results show that it captures more deep user search task information and performs as good as Query task clustering for query suggestion and opens the gates for URL recommendation framework which is input to search personalizing.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.