RANKING CLASSES OF SEARCH ENGINE RESULTS

Abstract

Ranking search results is an ongoing research topic in information retrieval. The traditional models are the \nvector space, probabilistic and language models, and more recently machine learning has been deployed in an \neffort to learn how to rank search results. Categorization of search results has also been studied as a means to \norganize the results, and hence to improve users search experience. However there is little research to-date on \nranking categories of results in comparison to ranking the results themselves. \n \nIn this paper, we propose a probabilistic ranking model that includes categories in addition to a ranked results \nlist, and derive six ranking methods from the model. These ranking methods utilize the following features: the \nclass probability distribution based on query classification, the lowest ranked document within each class and \nthe class size. \n \nAn empirical study was carried out to compare these methods with the traditional ranked-list approach in terms \nof rank positions of click-through documents and experimental results show that there is no simpler winner in \nall cases. Better performance is attained by class size or a combination of the class probability distribution of \nthe queries and the rank of the document with the lowest list rank within the class.