https://www.selleckchem.com/products/mln2480.html
This work presents an alternative method to represent documents based on LDA (Latent Dirichlet Allocation) and how it affects to classification algorithms, in comparison to common text representation. LDA assumes that each document deals with a set of predefined topics, which are distributions over an entire vocabulary. Our main objective is to use the probability of a document belonging to each topic to implement a new text representation model. This proposed technique is deployed as an extension of the Weka software as a new filter. T
Like
Comment
Share
