Nowadays, there're many abbreviations such as "you"-"u", "are"-"r", "please"-"plz", "thanks"-"tks", the abbreviated words are generally simple characters or not integrated words with no meaning. If using the indexing phrases to solve the polysemy problem, actually the context are also abbreviated words which would make it more confusing, such as "r u goin 2 c the film b4 tmr?" How can these words be automatically indexed?
No comments:
Post a Comment