Probabilistic Topic Modeling of Text
Topic modeling is one of the most upcoming research topic in information retrieval. After topic detection and tracking in large documents, academic researchers and practitioners tend to work with topic detection among social media data
This is required because social media data comes from different social media sites including facebook, twitter, PInterest etc.
Large amount of data remain unused. To use this data, first, there exist need to understand that what this data is all about. Any post or tweet (microblog) can be related to one or more topics.
It may be possible that there exist different probabilities for every tweet to get associated with various topics as per different probabilities respectively.
This can be done using Python or R programming languages. Also, GATE toolkit is very useful for the same.
Some of the topic modeling algorithms include Latent Dirichlet Allocation, Probabilistic Latent Semantic Analysis, Relational Topic Modeling, Dynamic Topic Modeling, Structural Topic Modeling etc..
Topic modeling may be supervised, unsupervised or semi-supervised. These may include textual modeling representation or graphical modeling representation. These are also called probabilistic topic modeling techniques
Comments
Post a Comment