Assigning Character Strings to Links for Construction of Concept Networks

Abstract

In this paper, we propose a method of extracting character strings from newspaper articles and adding them to the links of a word network to express node relationships. Attaching such strings to links enables us show the word relationships. When we evaluated the results in terms of the mean reciprocal rank (MRR), Top-1 accuracy rate, and Top-5 accuracy rate and considered answers with additional or missing information to be correct, the MRR was about 0.7, the Top-1 accuracy rate was about 0.6, and the Top-5 accuracy rate was about 0.9.