Title :
Application of data mining for identifying topics at the document level
Author :
Reza, Marifa Farzin ; Matin, Rizwana
Author_Institution :
Comput. Sci. & Eng., BRAC Univ., Dhaka, Bangladesh
Abstract :
Data mining techniques are very popular in modern days and are used in NLP (Natural Language Processing). It allows users to analyze data from many different perspectives, categorize it, and summarize the relationships identified. One of the techniques, clustering items to groups, has been very popular. We use this technique here to find different topics in a document. We aim to replicate previous results and empirically verify this measure to identify hypothetical topic boundaries.
Keywords :
data mining; document handling; natural language processing; pattern clustering; NLP; data mining; document level topic identification; hypothetical topic boundaries; item clustering; natural language processing; Clustering algorithms; Data mining; Natural language processing; Noise; Prediction algorithms; Speech; Unsupervised learning; Artificial intelligence; Data-mining; Natural language Processing (NLP); Unsupervised Learning;
Conference_Titel :
Informatics, Electronics & Vision (ICIEV), 2013 International Conference on
Conference_Location :
Dhaka
Print_ISBN :
978-1-4799-0397-9
DOI :
10.1109/ICIEV.2013.6572712
