Hi, I'm try to write my own analyzer for Naive Bayes classification. When I run the seq2sparse command and called -a MyANalyzer, The output shows the algorithim used my filter and analyzer:
### Using length filter ### Using stopword fileter ### Using my filter ### Using my keyword filter But if I check the dictionary-file-0 generated in this step, I still find all the tokens (not filtered!). Does that mean my analyzer didn't work, or the dictionary-file-0 just record all tokens but the algorithm actually use the filtered data? Thanks Mengfei Ren
