A Survey of Topic Modeling in Text Mining
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2015, Vol 6, Issue 1
Abstract
Topic models provide a convenient way to analyze large of unclassified text. A topic contains a cluster of words that frequently occur together. A topic modeling can connect words with similar meanings and distinguish between uses of words with multiple meanings. This paper provides two categories that can be under the field of topic modeling. First one discusses the area of methods of topic modeling, which has four methods that can be considerable under this category. These methods are Latent semantic analysis (LSA), Probabilistic latent semantic analysis (PLSA), Latent Dirichlet allocation (LDA), and Correlated topic model (CTM). The second category is called topic evolution models, which model topics by considering an important factor time. In the second category, different models are discussed, such as topic over time (TOT), dynamic topic models (DTM), multiscale topic tomography, dynamic topic correlation detection, detecting topic evolution in scientific literature, etc.
Authors and Affiliations
Rubayyi Alghamdi, Khalid Alfalqi
Keywords
Related Articles
- EP ID EP116571
- DOI 10.14569/IJACSA.2015.060121
- Views 132
- Downloads 3
An Efficient Deep Learning Model for Olive Diseases Detection
Worldwide, plant diseases adversely influence both the quality and quantity of crop production. Thus, the early detection of such diseases proves efficient in enhancing the crop quality and reducing the production loss....
Development of Talent Model based on Publication Performance using Apriori Technique
The main problem or challenge faced by Human Resource Management (HRM) is to recognize, develop and manage talent efficiently and effectively. This is because HRM is responsible for selecting the correct talent for suita...
Empirical Study of Segment Particle Swarm Optimization and Particle Swarm Optimization Algorithms
In this paper, the performance of segment particle swarm optimization (Se-PSO) algorithm was compared with that of original particle swarm optimization (PSO) algorithm. Four different benchmark functions of Sphere, Rosen...
Novel Adaptive Auto-Correction Technique for Enhanced Fingerprint Recognition
Fingerprints are the most used biometric trait in applications where high level of security is required. Fingerprint image may vary due to various environmental conditions like temperature, humidity, weather etc. Hence,...
Context-Sensitive Opinion Mining using Polarity Patterns
The growing of Web 2.0 has led to huge information is available. The analysis of this information can be very useful in various fields. In this regards, opinion mining and sentiment analysis are one of the most interesti...