Distributed LDA-based Topic Modeling and Topic Agglomeration in a Latent Space.
Gopi Chand Nutakki, Olfa Nasraoui, Behnoush Abdollahi, Mahsa Badami, Wenlong Sun · 2014
We describe the methodology that we followed to automatically extract topics corresponding to known events provided by the SNOW 2014 challenge in the context of the SocialSensor project. A data crawling tool and selected fil-tering terms were provided to all the teams. The crawled data was to be divided in 96 (15-minute) timeslots spanning a 24 hour pe-riod and participants were asked to produce a fixed number of topics for the selected times-lots. Our preliminary results are obtained us-ing a methodology that pulls strengths from several machine learning techniques, including Latent Dirichlet Allocation (LDA) for topic modeling and Non-negative Matrix Factoriza-tion (NMF) for automated hashtag annota-tion and for mapping the topics into a latent space where they become less fragmented and can be better related with one another. In ad-dition, we obtain improved topic quality when Copyright c © by the paper’s authors. Copying permitted only for private and academic purposes.