A Journal Paper Filtering Using the Profile Revised by Patent Document Information

Abstract

A paper filtering system that supports the effective collection of related journal papers is becoming important as the technological progress has been rapid. However, previous systems realized insufficient filtering effectiveness, because of lack of available information in abstract part of journal papers that is available to the public. In this paper, we propose a paper filtering method using the profile which is revised from not only relevant papers but also relevant patent documents by pseudo relevance feedback. By the evaluation for 15 examinees using 4,875 journal papers, it became clear that for all examinees the effectiveness of proposed method exceeds the conventional method, when the counterparts in patent document components for the journal paper's abstract is used, which became closer to practical use.