Reader Perspective Emotion Analysis in Text through Ensemble based Multi-Label Classification Framework

Abstract

Multiple emotions are often triggered in readers in response to text stimuli like news article. In this paper, we present a novel method for classifying news sentences into multiple emotion categories using an ensemble based multi-label classification technique called RAKEL. The emotion data consists of 1305 news sentences and the emotion classes considered are disgust, fear, happiness and sadness. Words are the most obvious choice as feature for emotion recognition. In addition to that we have introduced two novel feature sets: polarity of subject, verb and object of the sentences and semantic frames. Experiments concerning the comparison of features revealed that semantic frame feature combined with polarity based feature performs best in emotion classification. Experiments on feature selection over word and semantic frame features have been performed in order to handle feature sparseness problem. In both word and semantic frame feature, improvements in the overall performance have been observed after optimal feature selection.

Multi‐label learning: a review of the state of the art and ongoing research

“…It has been applied to many kinds of documents such as legal texts, web documents, news, research papers, narrative clinical text, patents, or aeronautics reports . Other related applications are document indexing, tag suggestion, e‐mail filtering, medical coding, query categorization, or the classification of news sentences into multiple emotion categories Multimedia .…”

“…22 Other related applications are document indexing, 23 tag suggestion, 24,25 e-mail filtering, 26 medical coding, 27 query categorization, 28 or the classification of news sentences into multiple emotion categories. 29,30 • Multimedia. MLL techniques have been applied to many types of resources such as images, videos, and sound.…”

Multi‐label emotion recognition of weblog sentence based on Bayesian networks

“…Lastly, sentences simply labeled a single emotion as positive or negative in traditional studies. While a sentence could have several emotion states, it could be assigned two or more emotion tags at the same time . In this paper, a multi‐label emotion topic model is constructed to recognize the complex emotions of weblog sentences on the basis of the Chinese emotion corpus Ren‐CECps with detailed emotional tags in each article.…”

Identifying subjective statements in news titles using a personal sense annotation framework

“…The data set we are considering in this section has been described (Strapparava & Mihalcea, ) and is widely used in research (Bhowmick, Basu, & Mitra, ; Bao et al., ). It consists of 1,250 news headlines from major newspapers such as BBC News, and from the Google News search engine.…”