You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently MeCab plugin just splits the input string and use surface of each morpheme as a feature.
It is better to support morpheme n-gram to extract the context of the natural language.
I'm thinking of adding a new "ngram" parameter (which is optional and defaults to 1) to the mecab plug-in configuration as follows:
For instance, when 本日は晴天 is given as an input, {"本日", "は", "晴天"} is extracted when "ngram" is 1, whereas {"本日|は", "は|晴天"} is extracted when "ngram" is 2.
The text was updated successfully, but these errors were encountered:
Currently MeCab plugin just splits the input string and use surface of each morpheme as a feature.
It is better to support morpheme n-gram to extract the context of the natural language.
I'm thinking of adding a new "ngram" parameter (which is optional and defaults to 1) to the mecab plug-in configuration as follows:
For instance, when
本日は晴天
is given as an input,{"本日", "は", "晴天"}
is extracted when "ngram" is 1, whereas{"本日|は", "は|晴天"}
is extracted when "ngram" is 2.The text was updated successfully, but these errors were encountered: