Hi Hive account@smcaterpillar, just another question: Do you plan on implementing Probabilistic latent semantic analysis (PLSA), or anything else, as an improvement to the LSA approach? I saw that LSA has a hard time with polysemy, and cannot detect different subtlety like different meanings for the same word when occurring in different phrases. Does PLSA would help with that? I am curious about that. I imagine computation time could become a issue thou... could you comment about that?

b.b.