Semantic Graph Based Term Expansion for Sentence-Level Sentiment Analysis

Abstract

The semantic orientation (also referred to as prior polarity) of a word plays an important role in automatic sentence-level sentiment analysis. Several approaches have been proposed wherein a lexicon of words marked with their polarities is exploited to infer the meaning of sentences. However, relying on prior word polarity may produce inaccurate decisions. This is because we may find negative-sentence sentiments that include words with positive prior polarities or vice versa. In this article, we propose an approach to sentence-level sentiment analysis that exploits knowledge encoded in heavy-weight semantic graphs to assist in discovering the meaning of a word in the context of the sentence where it appears. In this context, we build contextual semantic networks for indexing sentences and expand them with semantically/lexically-relevant terms in an attempt to disambiguate the meanings of word mentions in sentences. In order to verify the effectiveness of the proposed approach, we have developed a prototype system using a real-world dataset that contains 46830 sentiment sentences along with a gold-standard that comprises 10000 movie reviews that are labelled under five sentiment categories (very negative, negative, neutral, positive, very positive). Findings indicate that enriching the semantic graphs of sentiment sentences with NOUN-based synonyms and hypernyms has improved the overall quality of baseline sentiment analysis techniques.

A hybrid composite features based sentence level sentiment analyzer

“…On the other hand, other studies showed that coupling bigrams and trigrams based on extrinsic semantic resources provided better results than unigrams alone [2], [18]. As reported by Maree and Eleyat in [2], incorporating high-order ngrams such as trigrams that can be acquired based on external knowledge resources captures a sentence contextual information more precisely than other unigram or bigram-based tokenizers. Accordingly, and in light of these conclusions, it is crucial to experimentally evaluate the utilization of n-gram tokenization as part of the large-scale SA process.…”

“…Sources of sentiment sentences can vary on the Web. Twitter and other social media websites are among the most commonly referred to sources [2]. For experimental evaluation purposes, we use the wellknown internet movie database (IMDB) movie reviews dataset, which is publicly available at Kaggle.…”

“…As we discussed in the literature review section, there are several SA models that employ different feature extraction techniques. For instance, considering the n-gram tokenization process, choosing a specific size for n-gram tokens affects the overall's accuracy of the sentiment analyzer, especially when dealing with sentences contain negations, such as "not good" or "not very good" as reported in [2], [16]. In this context, we can see that using unigram features to process a sentence with "not good", will separate the word "good" from the negation word "not".…”

“…In an earlier study in 2004, Pang and Lee has showed that unigram features are more significant than bigrams when performing emotional classification for movie reviews [17]. On the other hand, other studies showed that coupling bigrams and trigrams based on extrinsic semantic resources provided better results than unigrams alone [2], [18]. As reported by Maree and Eleyat in [2], incorporating high-order ngrams such as trigrams that can be acquired based on external knowledge resources captures a sentence contextual information more precisely than other unigram or bigram-based tokenizers.…”

“…Sentiment analysis (SA) has become one of the most reliable tools for assisting organizations better understand the perception of their users about the products and services that they offer [1], [2]. Exploiting SA techniques in real-world application domains also serves individuals who are interested in learning more about the various customer perceptions on the products and/or services of interest.…”