Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10480606 | Physica A: Statistical Mechanics and its Applications | 2013 | 12 Pages |
Abstract
Word-clustering and entropic keyword detectors are compared in long and short texts. ⺠We propose two suitable metrics to evaluate the performance of keyword detectors. ⺠The word-clustering detector is improved to find relevant low-frequency words. ⺠The entropic detector gives reliable results only for suitable text partitions. ⺠The word clustering detector is successful both in long and short texts.
Related Topics
Physical Sciences and Engineering
Mathematics
Mathematical Physics
Authors
C. Carretero-Campos, P. Bernaola-Galván, A.V. Coronado, P. Carpena,