Skip to Main Content
Huge amount of unstructured data is available in the form of text documents. Ranking these text documents by considering their context will be very useful in information retrieval. We propose classification of abstracts by considering their context using Naïve Bayes classifier and Apriori association rule algorithm - i.e. Context Based Naive Bayesian and Apriori (CBNBA). In proposed approach, we initially classify the documents using Naïve Bayes. We find the context of an abstract by looking for associated terms which help us understand the focus of the abstract and interpret the information beyond simple keywords. The results indicate that context based classification increases accuracy of classification to great extent and in turn discovers different contexts of the documents. Further this approach can found to be very useful for applications beyond abstract classification where word speaks very little and lead to ambiguous state but context can lead you to right decision/classification.
Date of Conference: 5-7 Sept. 2012