Topic Categorization on Social Network Using Latent Dirichlet Allocation

Topic modelling is a powerful technique for analysis of large document collection. Topic modelling is used for finding hidden topic from the collection of document. In the twitter api, it is essential all the tweet documents are properly categorized. For automatically categorizing the twitter document topics The efficient detection is modelled by an LDA method for probabilistic model and for separation of words from the document. LDA is widely used to estimate the multinomial observation and each topic is categorized by a probabilistic distribution over the words. The multinomial distribution of the topics is regarded as the feature of the document. The proposed system resulted in an increase in accuracy for detection of the topic categorization.