Topic Modelling Twitter Data with Latent Dirichlet Allocation Method
- 1 October 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 386-390
Abstract
Twitter is a popular social media for every user to issue thoughts and emotional forms which are tweets, tweets that only have 140 characters with limitations to write in text. Twitter is one of the social media places to get information that is always up to date, tweets are categorized into big data because tweets are information that can be used as a source of data for research. Latent Dirichlet Allocation (LDA) as an algorithm that can process large text data (big data). In this study using the LDA method as an algorithm to produce topic modeling, each topic similarity, and visualization of topic clusters from the tweet data generated as many as 4 topics (Economic, Military, Sports, Technology) in Indonesian, where each topic has a number different tweets. The LDA method used in the processing of tweet data is successfully carried out and works optimally, in each topic extraction, topic modeling, generating index words that are in each topic cluster and computer visualization in the topic.LDA output shows optimal performance in the process of word indexing in Sport topics with 1260 tweets with an accuracy of 98% better than the LSI method in Topic Modeling.Keywords
This publication has 10 references indexed in Scilit:
- A REVIEW ON OVERLAPPING AND NON-OVERLAPPING COMMUNITY DETECTION ALGORITHMS FOR SOCIAL NETWORK ANALYTICSFar East Journal of Electronics and Communications, 2018
- RESEARCHGATE DATA ANALYSIS TO MEASURE THE STRENGTH OF INDONESIAN RESEARCHFar East Journal of Electronics and Communications, 2017
- Short text classification based on LDA topic model2017
- Probabilistic topic modelsCommunications of the ACM, 2012
- Document Clustering Using Concept Space and Cosine Similarity Measurement2009
- Clustering2008
- Incremental document clustering using cluster similarity histograms2004
- An introduction to latent semantic analysisDiscourse Processes, 1998
- Computational experience on four algorithms for the hard clustering problemPattern Recognition Letters, 1996
- Indexing by latent semantic analysisJournal of the American Society for Information Science, 1990