Concept Identification from Poems

Abstract

Emotion recognition has become an important topic in natural language processing. Usually words labeled with their emotion is the starting place to find the emotion of a text,but it is very essential that context must also be considered. It is not a simple task to capture the overall emotion of a Text, as words are mutually influence their emotion related interpretation.There is a lot of word-based dictionaries available for emotion recognition from the text. However, commonsense knowledge bases are very less. We used ConceptNet ( commonsense knowledge) for concept mining from poems. In this paper we proposed a concept identification method by using ConceptNet knowledge base to depict the concept of poems.This system gives a precision value of 71% and inter-rater agreement of 48%, depicting its moderate agreement with Human Expert interpretation.