A new feature selection based on comprehensive measurement

A new feature selection based on comprehensive measurement both in inter -category and intra-category for text categorization Presenter : Bo-Sheng Wang Authors : Jieming Yanga, b, Yuanning Liua, Xiaodong Zhua, *, Zhen Liua, c, Xiaoxu Zhanga IPM, 2012 1

Outlines • • • 2 Motivation Objectives Methodology Experiments Conclusions Comments

Motivation 3

Objectives • They proposed a new feature selection algorithm, named CMFS. – It comprehensively measures the significance of a term both in inter-category and intra-category. 4

Methodology 5

Methodology. Algorithm 6

Experiments • Classification algorithm – Naïve Bayes classifier – Support Vector Machines • Dataset – 20 -Newsgroups – Reuters-21578 – Web. KB • Preprocessing – Converted to lower case, remove punctuation, stop lists were use, no stemming was used. 7

Experiments • we measured the performance of the text categorization in terms of F 1 and Accuracy. 8

Experiments • 20 -Newsgroups 9

Experiments 10

Experiments • Reuters-21578 11

Experiments 12

Experiments • Web. KB 13

Experiments 14

Experiments • Statistical 15

Conclusions • According to above experiment, we can know CMFS better than other feature selection algorithm. 16

Comments • Advantages – It have high accuracy and performance. • Applications – Feature selection – Text categorization 17