An Empirical Study for Determining Relevant Features for Sentiment Summarization of Online Conversational Documents

A topic-enhanced word embedding for Twitter sentiment classification

“…Twitter sentiment classification, which aims at identifying the sentiment polarity of short and informal tweets, has attracted increasingly research interest in recent years [16,17,18,24,29,43,49,51]. Existing approaches are dominated by two mainstream lines.…”

“…Twitter sentiment classification, which aims to classify the sentiment polarity of a tweet as positive, neutral or negative, has been intensively researched in recent years [16,17,18,24,29,43,49]. Most work follows Pang et al (2002)' work, building a classifier based on annotated corpus with manually-designed sophisticated features [37].…”

Polarity classification using structure-based vector representations of text

“…In other work, text has been represented as a bag-of-opinions, where features denote occurrences of unique combinations of opinion-conveying words, amplifiers, and negators [46]. Other features capture the length of a text segment, and the extent to which it conveys opinions [2,20].…”

“…In the past decade, the Web has experienced an exponential growth into a network of more than 555 million Web sites, with over two billion users [1]. The Web has become an influential source of information with an increasing share of user-generated content, produced by many contributors [2]. This ubiquitous and ever-expanding usergenerated content ranges from (micro)blog posts to reviews.…”

Word2Sent: A new learning sentiment‐embedding model with low dimension for sentence level sentiment classification

“…However, lexicon‐based approaches have many issues such as the need for powerful linguistic resources which are not always available. Other machine learning algorithms are based on manual feature techniques 2,14 that take up a lot of time. Despite the progress made by these methods, extracting useful features and choosing the appropriate algorithm are very important in sentiment analysis applications.…”