Incremental Algorithms for Effective and Efficient Query Recommendation

Abstract

Abstract. Query recommender systems give users hints on possible interesting queries relative to their information needs. Most query recommenders are based on static knowledge models built on the basis of past user behaviors recorded in query logs. These models should be periodically updated, or rebuilt from scratch, to keep up with the possible variations in the interests of users. We study query recommender algorithms that generate suggestions on the basis of models that are updated continuously, each time a new query is submitted. We extend two stateof-the-art query recommendation algorithms and evaluate the effects of continuous model updates on their effectiveness and efficiency. Tests conducted on an actual query log show that contrasting model aging by continuously updating the recommendation model is a viable and effective solution.

An Efficient Incremental Query Recommendation based Query Ranking Model

“…The incrementally updating mechanism is not provided in the existing query recommendation techniques. These models must be re built, by incorporating the new log data [3]. Practically, it is known that the building of existing models is very time consuming.…”

A New Query Suggestion Algorithm for Taxonomy-based Search Engines

“…So it is possible to classify the systems in context aware, like (Cao et al, 2008), and non context aware. We can also say an algorithm is incremental or non incremental, depending on whether it modifies its internal structure as it is given a new query or it has a fixed set-up and may be only queried (Broccolo et al, 2010).…”

“…Following the classification given by (Broccolo et al, 2010), we can consider our algorithm as an incremental session-based non context-aware approach.…”