Base | L | By unconditional probability | By 1st fragment | By 2nd fragment | By mean probability | Method 1 | Method 2 | Method 3 | Method 4 | NFIM | Advantage of NFIM (%) (37) |
All words | 1 | 0.09457 | 0.08879 | 0.09018 | 0.08228 | 0.08170 | 0.07842 | 0.08182 | 0.08117 | 0.08090 | −3.05 |
Without words “всё” and “это” | 1 | 0.095 | 0.08781 | 0.08957 | 0.08204 | 0.08112 | 0.07844 | 0.08079 | 0.08021 | 0.0779 | 0.7 |