Annotation, Modelling and Analysis of Fine-Grained Emotions on a
            Stance and Sentiment Detection Corpus

Schuff, Hendrik; Barnes, Jeremy; Mohme, Julian; Padó, Sebastian; Klinger, Roman

doi:10.18653/v1/w17-5203

Cited by 71 publications

(70 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We compare the performance of our proposed system with the state-of-the-art systems of SemEval 2016 Task 6 and the systems of [16]. Experimental results show that the proposed system improves the existing state-of-the-art systems for sentiment and emotion analysis.…”

Section: Results and Analysismentioning

confidence: 97%

See 1 more Smart Citation

Emotion helps Sentiment: A Multi-task Model for Sentiment and Emotion Analysis

Abhishek

Ekbal

Kawahra

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

In this paper, we propose a two-layered multi-task attention based neural network that performs sentiment analysis through emotion analysis. The proposed approach is based on Bidirectional Long Short-Term Memory and uses Distributional Thesaurus as a source of external knowledge to improve the sentiment and emotion prediction. The proposed system has two levels of attention to hierarchically build a meaningful representation. We evaluate our system on the benchmark dataset of SemEval 2016 Task 6 and also compare it with the state-of-theart systems on Stance Sentiment Emotion Corpus. Experimental results show that the proposed system improves the performance of sentiment analysis by 3.2 F-score points on SemEval 2016 Task 6 dataset. Our network also boosts the performance of emotion analysis by 5 F-score points on Stance Sentiment Emotion Corpus.

show abstract

Section: Results and Analysismentioning

confidence: 97%

“…Sigmoid cross-entropy was used as the loss function. F1-score was reported for the sentiment analysis [8] and precision, recall and F1-score were used as the evaluation metric for emotion analysis [16]. Therefore, we report the F1-score for sentiment and precision, recall and F1-score for emotion analysis.…”

Section: Implementation Detailsmentioning

confidence: 99%

Emotion helps Sentiment: A Multi-task Model for Sentiment and Emotion Analysis

Abhishek

Ekbal

Kawahra

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

show abstract

“…The statistics of annotated tweets are presented in Table II. [13]. We have prepared two datasets of annotated tweets considering majority votes and all votes.…”

Section: Dataset Collection and Annotationmentioning

confidence: 99%

Multilabel Classification for Emotion Analysis of Multilingual Tweets

Gohil¹,

Patel²

2019

IJITEE

View full text Add to dashboard Cite

Emotion Analysis of text targets to detect and recognize types of feelings expressed in text. Emotion analysis is successor of Sentiment analysis. The latter does coarse-level analysis and classify the text into positive and negative categories while former does fine-grain analysis and classify text in specific emotion categories like happy, surprise, angry. Analysis of text at fine-level provides deeper insight compared to coarse-level analysis. In this paper, tweets are classified in discrete eight basic emotions namely joy, trust, fear, surprise, sadness, anticipation, anger, disgust specified in Plutchik’s wheel of emotions [1]. Tweets for three languages collected out of which one is English language and rest two are Indian languages namely Gujarati and Hindi. The collected tweets are related to Indian politics and are annotated manually. Supervised Learning and Hybrid approach are used for classification of tweets. Supervised learning uses tf-idf as features while hybrid approach uses primary and secondary features. Primary features are generated using tf-idf weighting and two different algorithms of feature generation are proposed which generate secondary features using SenticNet resource. Multilabel classification is performed to classify tweets in emotion categories. Results of experiments show effectiveness of hybrid approach.

show abstract

“…The first manually-annotated corpus of tweets for emotion analysis made publicly available was provided by [19], followed by a larger set with a focus on emotion intensity prediction [20]. The corpus by [21] provides multiple annotations of each instance and analyzes interactions between classes. It is a re-annotation of a SemEval corpus for stance detection [22].…”

Section: Background and Related Work A Emotion Analysismentioning

confidence: 99%

An Empirical Analysis of the Role of Amplifiers, Downtoners, and Negations in Emotion Classification in Microblogs

Strohm¹,

Klinger

2018

2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA)

Self Cite

View full text Add to dashboard Cite

The effect of amplifiers, downtoners, and negations has been studied in general and particularly in the context of sentiment analysis. However, there is only limited work which aims at transferring the results and methods to discrete classes of emotions, e. g., joy, anger, fear, sadness, surprise, and disgust. For instance, it is not straight-forward to interpret which emotion the phrase "not happy" expresses. With this paper, we aim at obtaining a better understanding of such modifiers in the context of emotion-bearing words and their impact on document-level emotion classification, namely, microposts on Twitter. We select an appropriate scope detection method for modifiers of emotion words, incorporate it in a document-level emotion classification model as additional bag of words and show that this approach improves the performance of emotion classification. In addition, we build a term weighting approach based on the different modifiers into a lexical model for the analysis of the semantics of modifiers and their impact on emotion meaning. We show that amplifiers separate emotions expressed with an emotionbearing word more clearly from other secondary connotations. Downtoners have the opposite effect. In addition, we discuss the meaning of negations of emotion-bearing words. For instance we show empirically that "not happy" is closer to sadness than to anger and that fear-expressing words in the scope of downtoners often express surprise.

show abstract

Annotation, Modelling and Analysis of Fine-Grained Emotions on a Stance and Sentiment Detection Corpus

Cited by 71 publications

References 38 publications

Emotion helps Sentiment: A Multi-task Model for Sentiment and Emotion Analysis

Emotion helps Sentiment: A Multi-task Model for Sentiment and Emotion Analysis

Multilabel Classification for Emotion Analysis of Multilingual Tweets

An Empirical Analysis of the Role of Amplifiers, Downtoners, and Negations in Emotion Classification in Microblogs

Contact Info

Product

Resources

About