nlp - Using pre-defined topics in Mallet -
nlp - Using pre-defined topics in Mallet -
i'm looking utilize mallet classify different documents topics have defined. know mallet first determine topics, classify documents want skip first step because have list of topics words associated them. there way utilize pre-defined topic lists have created classify documents mallet?
any guidance appreciated. thanks!
if you're doing unsupervised learning (without training examples, i.e. docs each topic), cannot trivially set topics. point training algorithm not know docs in advance. tries separate/distribute them, based on features provide.
if you're doing supervised learning, topics classes , have documents each class. algorithm tries larn features important each class. in mallet should utilize classification module.
there fancy topic modelling ideas, incorporate / skew topic distributions according specific keywords, don't think that's possible mallet.
nlp topic-modeling mallet
Comments
Post a Comment