A comparative study of feature selection and machine learning techniques for sentiment analysis

A Cross-Domain Analysis using Morphological Sentence Pattern Approach for Extracting Aspect-based Lexicon

“…The purpose of sentiment analysis is extracting opinions or emotional states regarding certain topics such as events, products, entertainers, politicians and movies from the textual data to find people"s interests and thoughts [3][4][5]. Especially, aspect-based sentiment analysis is one of advanced approach of lexicon based sentiment analysis.…”

“…Despite the demands of sentiment analysis methods for analyzing social media data, fundamental challenges still remain, because user-generated online textual data is unstructured, unlabeled, and noisy to be analyzed accurately. Especially, building lexicon usually needs human-coding efforts because the lexicon affects a quality of analysis in the lexicon based sentiment analysis approach [3][4][5].…”

A Comparative Study of Combined Feature Selection Methods for Arabic Text Classification

“…A high weight is assigned to a feature if it differentiates between instances from different classes and has the same value for instances of the same class. Specifically, it tries to find a best estimate from the following probabilities to allocate as the weight for each term feature f (Sharma and Dey, 2012): …”

Is News Sentiment More than Just Noise?

“…Dictionary-based approaches produce reliable results by counting the frequency of pre-defined negative and positive words from a given dictionary. Machine learning methods (e. g. Antweiler and Frank, 2004;Li, 2010;Mittermayer and Knolmayer, 2006a;Schumaker and Chen, 2009) represent a variety of methods, but may be subject to overfitting (Sharma and Dey, 2012). A remedy may originate from regularization methods that utilize variable selection to generate domain-dependent dictionaries.…”