User Search Goals Evaluation with Feedback Sessions
V. Febna, Anish Abraham · Procedia Technology · 2016
In today's e-world, search engines play a vital role in retrieving and organizing relevant data for various purposes. Different methods are used to find user search goals. Personalization is the process of finding exact needs of a user using different representations and machine learning techniques. These methods exploit feedback sessions and bipartite graphs, along with machine learning techniques such as clustering, classification and Apriori algorithms. This paper proposes a variant of feedback session method for inferring user search goals, where bag of words approach is employed for representation. K-Medoid clustering algorithm is used to derive the cluster for the keywords entered by the user. The performance improvement can be evaluated by using evaluation measures like Average Precision (AP), Voted Average Precision (VAP) and Classified Average Precision (CAP).