Generating a Word-Emotion Lexicon from #Emotional Tweets

Abstract

Research in emotion analysis of text suggest that emotion lexicon based features are superior to corpus based n-gram features. However the static nature of the general purpose emotion lexicons make them less suited to social media analysis, where the need to adopt to changes in vocabulary usage and context is crucial. In this paper we propose a set of methods to extract a word-emotion lexicon automatically from an emotion labelled corpus of tweets. Our results confirm that the features derived from these lexicons outperform the standard Bag-of-words features when applied to an emotion classification task. Furthermore, a comparative analysis with both manually crafted lexicons and a state-of-the-art lexicon generated using Point-Wise Mutual Information, show that the lexicons generated from the proposed methods lead to significantly better classification performance.

Emotion-Corpus Guided Lexicons for Sentiment Analysis on Twitter

“…Our contributions in this paper are as follows: 1. We propose two different methods to generate sentiment lexicons from a corpus of emotion-labelled tweets by combining our prior work on domain-specific emotion lexicon generation [1,2], with the emotion-sentiment mapping presented in Psychology (see figure 1) [4]; and 2. We comparatively evaluate the quality of the proposed sentiment lexicons, and the standard sentiment lexicons found in literature through different sentiment analysis tasks: sentiment intensity prediction and sentiment classification on benchmark Twitter data sets.…”

Lexicon based feature extraction for emotion text classification

“…In the following section we briefly explain our proposed lexicon generation method. Further details about our proposed DSEL generation can be found in [23,24].…”

“…Further crowd-annotated emotional news articles 2 were leveraged for lexicon generation, by combining the document-frequency distributions of words and the emotion distributions over documents [21,37]. In our prior work [23,24], we jointly modelled the emotionality and neutrality of words using a unigram mixture model (UMM) to learn DSELs from labelled and weakly-labelled emotion documents.…”

“…This is different from our work in [23] since it extracts features beyond simple word counts using the knowledge of a DSEL;…”

Neural Induction of a Lexicon for Fast and Interpretable Stance Classification

“…CPBL learning techniques are the baseline against which we evaluate our lexicon induction algorithm. They operate by computing the conditional probability of observing each lexicon entry under each class [1]. The value for each pair (t, c) where term t ∈ T and class c ∈ C is computed as indicated in equation 1.…”