Sentiment strength detection in short informal text

Abstract

A huge number of informal messages are posted every day in social network sites, blogs and discussion forums. Emotions seem to be frequently important in these texts for expressing friendship, showing social support or as part of online arguments. Algorithms to identify sentiment and sentiment strength are needed to help understand the role of emotion in this informal communication and also to identify inappropriate or anomalous affective utterances, potentially associated with threatening behaviour to the self or others. Nevertheless, existing sentiment detection algorithms tend to be commercially-oriented, designed to identify opinions about products rather than user behaviours. This article partly fills this gap with a new algorithm, SentiStrength, to extract sentiment strength from informal English text, using new methods to exploit the de-facto grammars and spelling styles of cyberspace. Applied to MySpace comments and with a lookup table of term sentiment strengths optimised by machine learning, SentiStrength is able to predict positive emotion with 60.6% accuracy and negative emotion with 72.8% accuracy, both based upon strength scales of 1-5. The former, but not the latter, is better than baseline and a wide range of general machine learning approaches.

A Blackboard System for Generating Poetry

“…The valence of the text is calculated by using the SentiStrength tool, which estimates the negative and positive sentiment strength in sentences based on the sentiment word strength list containing words with positive and negative emotional values evaluated by human judges [42]. However, as we would like to incorporate an independent emotional intelligence in our system, we introduce another parameter called the optimism rate, which can be set by the user at the beginning of the algorithm or chosen randomly.…”

SentiCircles for Contextual and Conceptual Semantic Sentiment Analysis of Twitter

“…Instead of using training data to learn sentiment, lexicon-based methods rely on pre-built dictionaries of words with associated sentiment orientations [20], such as SentiWordNet [3] or the MPQA subjectivity lexicon [26]. Thelwall et al [23,22] proposed SentiStrength; a lexicon-based method for sentiment detection on the social web. This method overcomes the common problem of ill-formed language on Twitter and the like, by applying several lexical rules, such as the existence of emoticons, intensifiers, negation and booster words (e.g., absolutely, extremely).…”

“…One limitation of lexicons is their static sentiment values of terms, regardless of their contexts. Although authors in [23] proposed an algorithm to update the sentiment strength assigned to terms in a lexicon, this algorithm required training from manually annotated corpora.…”

“…SenticNet was proved valuable for sentiment detection in conventional text (e.g., product reviews) [11]. Unlike SentiStrength [23], SenticNet is not tailored for Twitter and the like. Although conceptual semantic approaches have been shown to outperform purely syntactical approaches [7], they are usually limited by the scope of their underlying knowledge bases, which is especially problematic when processing general Twitter streams, with their rapid semiotic evolution and language deformations.…”

“…The aim is to investigate the ability of SentiCircles in updating these contextfree prior sentiment scores based on the contextual semantics extracted from different tweets corpora. We selected three state-of-art lexicons for this study: (i) the SentiWordNet lexicon [3], (ii) the MPQA subjectivity lexicon [26] and, (iii) Thelwall-Lexicon [23,22].…”