Missing words in similarity evaluation regarding minfreq
Closed this issue · 0 comments
nicolasdugue commented
Describe the bug
The number of missing words is increasing when min_freq is increased. Meanwhile, it should not increase thanks to the exception list.
To Reproduce
- Evaluate similarity with a corpus generated with a given minfreq
- Increase minfreq
Expected behavior
The number of missing word should not increase. This number is supposed to be the same for a given corpus.