Social Networking sites have become popular and common places for sharing wide range of emotions through short texts. These emotions include happiness, sadness, anxiety, fear, etc. Analyzing short texts helps in identifying the sentiment expressed by the crowd. Sentiment Analysis on IMDb movie reviews identifies the overall sentiment or opinion expressed by a reviewer towards a movie. Many researchers are working on pruning the sentiment analysis model that clearly identifies and distinguishes between a positive review and a negative review. In the proposed work, we show that the use of Hybrid features obtained by concatenating Machine Learning features (TF, TF-IDF) with Lexicon features (Positive-Negative word count, Connotation) gives better results both in terms of accuracy and complexity when tested against classifiers like SVM, Naïve Bayes, KNN and Maximum Entropy. The proposed model clearly differentiates between a positive review and negative review. Since understanding the context of the reviews plays an important role in classification, using hybrid features helps in capturing the context of the movie reviews and hence increases the accuracy of classification.

No abstract

Abstract In recent internet era, micro-blogging sites produce enormous amount of short textual information, which appears in the form of opinions or sentiments of users. Sentiment analysis is a challenging task in short text, due to use of formal language, misspellings, and shortened forms of words, which leads to high dimensionality and sparsity. In order to deal with these challenges, this paper proposes a novel, simple, and yet effective feature selection method, to select frequently distributed features related to each class. In this paper, the feature selection method is based on class-wise information, to identify the relevant feature related to each class. We evaluate the proposed feature selection method by comparing with existing feature selection methods like chi-square ( χ2), entropy, information gain, and mutual information. The performances are evaluated using classification accuracy obtained from support vector machine, K nearest neighbors, and random forest classifiers on two publically available datasets viz., Stanford Twitter dataset and Ravikiran Janardhana dataset. In order to demonstrate the effectiveness of the proposed feature selection method, we conducted extensive experimentation by selecting different feature sets. The proposed feature selection method outperforms the existing feature selection methods in terms of classification accuracy on the Stanford Twitter dataset. Similarly, the proposed method performs competently equally in terms of classification accuracy compared to other feature selection methods in most of the feature subsets on Ravikiran Janardhana dataset.

No abstract

With the advent of micro-blogging sites, users are pioneer in expressing their sentiments and emotions on global issues through text. Automatic detection and classification of sentiments like sarcastic or ironic content in microblogging reviews is a challenging task. It requires a system that manages some kind of knowledge to interpret the sentiment expressed in text. The available approaches are quite limited in their capabilities and scope to detect ironic utterances present in the text. In this regards, the paper propose feature fusion to provide knowledge to the system by alternative sets of features obtained using linguistic and content based text features. The proposed work extracts five sets of linguistic features and fuses with features selected using two stages of a feature selection method. In order to demonstrate the effectiveness of the proposed method, we conduct extensive experimentation by selecting different feature subsets. The performances of the proposed method are evaluated using Support Vector Machine (SVM), Logistic Regression (LR), Random Forest (RF), Decision Tree (DT) and ensemble classifiers. The experimental result shows the proposed approach significantly out-performs the conventional methods.