Keyword Extraction for Medium-Sized Documents Using Corpus-Based Contextual Semantic Smoothing

Table 2

Different combinations of POS.

S#		POS
1	Top-,	NO and AD only
2		NO, AD, and F only
3		NO, AD, and I only
4		NO, AD, I, and F only
5		NO, AD, and V only
6		NO, AD, I, V, NU, G, and AG only
7		All POS except MV

Bold values represent the best results achieved in terms of F-measure.