CROWDSOURCING A WORD–EMOTION ASSOCIATION LEXICON

In recent years, with the rapid development and wide application of the Internet, it has become the main place for the generation and dissemination of public opinion. To grasp the information of network public opinion in a timely and comprehensive way can not only effectively prevent sudden network malignant events but also provide a reference for the scientific and democratic decision-making of government departments. Therefore, in view of the practical application needs, this article studies the emotional characteristics and the evolution of public opinion over time based on the emotional feature words of network public opinion participants. Firstly, the positive and negative emotional lexicon of HowNet emotional dictionary is used, and the commonly used emotional lexicon and expression symbols are added to the lexicon. At the same time, the polarity annotation method of Chinese emotional lexicon ontology is used to construct the emotional lexicon of this article. Secondly, considering other emotional polarity characteristics in the dictionary, an emotional tendency analysis model is proposed. In this article, emotional analysis is applied to the evolution analysis of network public opinion, and the change of network public opinion characteristics with time series is obtained. The simulation results show that the emotional dictionary constructed in this article and the proposed model of emotional orientation analysis can effectively analyze the emotional characteristics of network public opinion participants and apply emotional analysis to the evolution analysis of network public opinion, which can get the change of emotional characteristics of public opinion participants with time series.

The Sociology of Emotions: Feminist, Cultural and Sociological Perspectives brings together conceptual and theoretical analysis drawing on feminist, cultural and sociological perspectives. The book explores the ‘language of emotions’, looking at macro and micro framing of emotions, including: emotions in modernity; emotions in late modernity; emotional labour and emotional capital; positive and negative emotions; masculinity and emotions; and love, intimacy and emotions. The book studies both positive and negative emotions including happiness, anger, fear, love, friendship, sadness, depression, shame and loneliness among others. The study of emotions is a relevant area of study for a more reflexive understanding of emotions. The book considers three dimensions in the analysis of emotions: the first considers conceptualizing emotions; the second considers theorizing emotions and the third, analyzing emotions. In conceptualizing emotions, the book explores the language of emotions looking at macro and micro framing of emotions. The theorizing of emotions considers both classical and contemporary social theories in understanding the emotions. Finally, the book focuses on analyzing emotions and includes empirical and conceptual research, including analysis of positive and negative emotions.

There is plenty of evidence that emotion analysis has many valuable applications. In this study a blog emotion corpus is constructed for Chinese emotional expression analysis. This corpus contains manual annotation of eight emotional categories (expect, joy, love, surprise, anxiety, sorrow, angry and hate), emotion intensity, emotion holder/target, emotional word/phrase, degree word, negative word, conjunction, rhetoric, punctuation and other linguistic expressions that indicate emotion. Annotation agreement analyses for emotion classes and emotional words and phrases are described. Then, using this corpus, we explore emotion expressions in Chinese and present the analyses on them.

<p>Emotion analysis, a recent sub discipline at the crossroads of information retrieval and computational linguistics is becoming increasingly important from application viewpoints of affective computing.Emotion is crucial to identify as it is not open to any objective observation or verification. In this paper, emotion analysis on blog texts has been carried out for a less privileged language, Telugu and the same system has been applied on the English SemEval 2007 affect sensing corpus containing only news headlines. A set of six emotion tags, namely, happy ( ), sad ( ), anger ( ), fear ( ), surprise ( )and disgust ( ), have been selected towards this emotion detection task for reliable and semi-automatic annotation of blog and news data. Conditional Random Field (CRF) based classifier has been applied for recognizing six basic emotion tags for different words of a sentence. The classifier accuracy has been improved by arranging an equal distribution of emotional tags and non-emotional tag. A score based technique has been adopted to calculate and assign tag weights to each of the six emotion tags. A sense based scoring strategy has been applied to identify sentence level emotion scores for the six emotion tags based on the acquired word level emotion tags. Sentence level emotion tagging has been carried out based on the maximum obtained sentence level emotion scores. Evaluation has been conducted for each emotion class separately on 200 test sentences from each of the Telugu blog and English news data. The system has resulted accuracies of 69.82% and 71.06% for happy, 70.24% and 66.42% for sad, 65.73% and 64.27% for anger, 76.01% and 69.90% for disgust, 72.19% and 73.59% for fear and 70.54% and 66.64% for surprise emotion classes on blog and news test data respectively.<br>  </p>

Tourist attractions need to optimize and upgrade their tourist services and tourist experience, according to the trendy topics on the Internet and the we media. This calls for objective evaluation of tourist experience, and accurately depiction of tourist emotions upon looking at tourism landscape images (TLIs). However, most of the existing methods for image emotional analysis cannot overcome the semantic gap, or handle an extraordinarily large image set. To solve the problems, this paper implements emotional analysis and annotation of TLIs based on tourist experience Firstly, the flow of tourist experience evaluation was expounded, and a model was constructed to evaluate tourist experience. Next, the forms of feature-based semantic information were specified for TLIs, and the emotional features were calculated for such images. After that, a semantic selection model was established to generate the emotional feature subsets of TLIs. Finally, the proposed model was verified through experiments on image emotional classification and annotation, and the relevant results were analyzed in details.

Emotional analysis techniques for text are very useful for analyzing public opinions on social media, and user reputation and opinions about products and services on electronic commerce sites. The Chinese Emotional Word Dictionary (CEWD) was created and has a very rich vocabulary for the fine-grained (11 emotion categories) emotional analysis of Chinese text. Additionally, the Chinese Emotional Expression Analysis System (CEEAS) that uses CEWD was developed to analyze emotions in Chinese text relatively accurately and quickly. In this study, we apply CEEAS to the emotional analysis of modern Chinese literature to investigate the characteristics of the appearance of emotional categories in modern Chinese literary works. Then, we apply CEEAS to modern translations of ancient Chinese literary works to perform emotional analysis, and compare the appearance characteristics of emotional expressions with those of modern literary works. Furthermore, we report on the changes in the appearance characteristics of emotional expressions between modern Japanese novels and their Chinese translations. We are convinced that literary research using the emotional analysis techniques proposed in this study will be useful as a new digital approach to literary research.

Since the celebration of the 100th anniversary of Darwin's The Language of the Emotions in Man and Animals (1872), emotionology has become a respectable and even thriving research domain again. The domain of human emotions is most important for mankind, emotions being right in the center of our daily lives and interests. A key-role in the interdisciplinary scientific debate about emotions has now been accorded to the study of the language of emotions. The present volume offers a new approach to the study of the language of emotions insofar as it presents theories from very different perspectives. It encompasses studies by scholars from diverse disciplines such as linguistics, sociology, and psychology. The topics of the contributions also cover a range of special fields of interest in four major sections. In a first section, a discussion of theoretical issues in the analysis of emotions is presented. The conceptualization of emotions in specific cultures is analyzed in section 2. Section 3 takes a different inroad into the language of emotions by looking at developmental approaches giving evidence of the fact that the acquisition of the language of emotions is a social achievement that simultaneously determines our experience of these emotions. Section 4 is devoted to emotional language in action, that is, the contributions focus upon different types of texts and analyze how emotions are referred to and expressed in discourse.

As a tool to study people’s views and opinions on things and events around them, sentiment analysis is widely used in the analysis and processing of mass evaluation information. Traditional emotion analysis is generally based on emotion dictionary, but the construction of emotion dictionary needs a lot of artificial time, and in different application fields need to establish different emotion dictionaries. Meanwhile, emotion dictionary can’t contain the semantic information of words and it also ignore the role of non-emotional words in the expression of emotion. This paper proposes a new emotion analysis algorithm (CBOW-PE) based on word embedding and part of speech. First, in the stage of pre-processed experiment, we use part of speech to preprocess the experimental data, fully consider the role of non-emotional words on emotion analysis. And then we add emotional information to assist word vector training, so that the word embedding of words related to emotion has both semantic information and emotional information. Finally, this paper makes a lot of experimental analysis and comparison of this paper and its related algorithms. The results show that the algorithm is effective and efficient in Chinese emotion analysis.

This paper analyzes selected examples of uses of argumentation tactics that exploit emotive language, many of them criticized as deceptive and even fallacious by classical and recent sources, including current informal logic textbooks. The analysis is based on six argumentation schemes, and an account of the dialectical setting in which these schemes are used. The three conclusions are (1) that such uses of emotive language are often reasonable and necessary in argumentation based on values, (2) but that they are defeasible, and hence need to be seen as open to critical questioning (3) and that when they are used fallaciously, it is because they interfere with critical questioning or conceal the need for it. The analysis furnishes criteria for distinguishing between arguments based on the use of emotive words that are reasonable tools of persuasion, and those that are fallacious tactics used to conceal and distort information.

Los campos de Análisis de Sentimientos y Emociones son prominentes en el Procesado de Lenguaje Natural (NLP, en inglés). Modelar emociones requiere de un número de técnicas que, en la mayoría de los casos, se comparten con otras áreas del NLP. Es este contexto, esta tesis aborda el desarrollo de métodos de aprendizaje automático novedosos a través de la combinación de características superficiales y profundas en Análisis de Sentimientos y Emociones. Tras observar los resultados obtenidos, los métodos desarrollados han sido adaptados a otras áreas del NLP, transfiriendo así los resultados obtenidos a campos que se pueden beneficiar en gran manera de dichas técnicas. De esta manera, hemos diseñado una taxonomía que permite clasificar diferentes tipos de técnicas en Análisis de Sentimientos, según las características y combinaciones usadas. Esta taxonomía nos ha permitido desarrollar varios modelos de aprendizaje automático que combinan características superficiales y profundas, además de un variedad de modelos de aprendizaje, sobre todo las técnicas de aprendizaje profundo. Estos modelos han sido evaluados mediante en entornos de Análisis de Sentimientos, a niveles de documento y aspecto. A continuación, hemos contribuido con una técnica para la generación de léxicos de sentimientos específicos del dominio, usando el algoritmo de retro-propagación. En una línea similar de trabajo, también hemos extendido un método para la generación de léxicos, obteniendo un recurso bilingüe anotado en Inglés e Italiano. La contribución central de esta tesis es un modelo que explota la similitud semántica en representaciones distribuidas a través de léxicos, extrayendo características mediante el uso de word embeddings y un léxico de sentimientos: el modelo SIMON. Hemos observado que SIMON ha mostrado resultado positivos en la evaluación experimental, y ofrece una alternativa potente a las técnicas clásicas de uso de léxicos. A la luz de esta última contribución, esta tesis estudia la adaptación de los métodos desarrollados a otros campos. Más concretamente, esta tesis ha contribuido a la (a) detección de radicalización, donde hemos adaptado el modelo SIMON, combinándolo con un método de extracción de características emotivas; y a la (b) estimación de valores morales, en la que se ha usado SIMON para emplear un nuevo léxico que hemos generado. ----------ABSTRACT---------- Sentiment and Emotion Analysis are prominent fields in Natural Language Processing (NLP) and have contributed to its progress. Modeling emotions requires a number of techniques and methods that, in most of the cases, are shared with other NLP areas. In such a context, this thesis addresses the development of novel machine learning methods through the combination of both surface and deep features for Sentiment and Emotion Analysis. Observing the results obtained, the developed methods have been adapted to other NLP areas, transferring the obtained results to fields that can largely benefit from such novel techniques. In this way, we have designed a taxonomy that classifies different approaches in Sentiment Analysis, attending to the features and combinations used. This taxonomy has allowed us to develop several machine learning models that combine both surface and deep features, as well as a variety of learning models, with an especial focus on deep learning approaches. These models have been evaluated in document and aspect-level Sentiment Analysis frameworks. Following, we contributed with a novel technique for generating domain-specific sentiment lexicons through the backpropagation algorithm. In a similar line of work, this thesis also extends a method for generating emotion lexicons, obtaining a bilingual resource with emotion annotations in both English and Italian. The thesis' core contribution is a similarity-based perspective on lexicons that extracts features exploiting both a word embedding model and a sentiment lexicon: the SIMilaritybased sentiment projectiON (SIMON) model. We have observed that SIMON has shown positive results in the experimental evaluation, offering a compelling alternative to classical lexicon usage techniques. In light of this last contribution, this thesis studies the adaptability of methods that are developed in the context of Sentiment and Emotion Analysis to other fields. More concretely, this thesis has contributed to radicalization detection and moral value estimation. In the area of radicalization detection, we have adapted the SIMON model, combining it with an emotion-driven feature extraction method. Similarly, for moral value estimation, SIMON has been used to exploit a novel lexicon we have generated.

In this article, we present a comprehensive study aimed at computing semantic relatedness of word pairs. We analyze the performance of a large number of semantic relatedness measures proposed in the literature with respect to different experimental conditions, such as (i) the datasets employed, (ii) the language (English or German), (iii) the underlying knowledge source, and (iv) the evaluation task (computing scores of semantic relatedness, ranking word pairs, solving word choice problems). To our knowledge, this study is the first to systematically analyze semantic relatedness on a large number of datasets with different properties, while emphasizing the role of the knowledge source compiled either by the ‘wisdom of linguists’ (i.e., classical wordnets) or by the ‘wisdom of crowds’ (i.e., collaboratively constructed knowledge sources like Wikipedia).The article discusses benefits and drawbacks of different approaches to evaluating semantic relatedness. We show that results should be interpreted carefully to evaluate particular aspects of semantic relatedness. For the first time, we employ a vector based measure of semantic relatedness, relying on a concept space built from documents, to the first paragraph of Wikipedia articles, to English WordNet glosses, and to GermaNet based pseudo glosses. Contrary to previous research (Strube and Ponzetto 2006; Gabrilovich and Markovitch 2007; Zeschet al. 2007), we find that ‘wisdom of crowds’ based resources are not superior to ‘wisdom of linguists’ based resources. We also find that using the first paragraph of a Wikipedia article as opposed to the whole article leads to better precision, but decreases recall. Finally, we present two systems that were developed to aid the experiments presented herein and are freely available1for research purposes: (i) DEXTRACT, a software to semi-automatically construct corpus-driven semantic relatedness datasets, and (ii) JWPL, a Java-based high-performance Wikipedia Application Programming Interface (API) for building natural language processing (NLP) applications.

The paper investigates the relation between emotions and feedback facial expressions in video and audio recorded Danish dyadic first encounters. In particular, we train a classifier on the manual annotations of the corpus in order to investigate to which extent the encoding of emotions contribute to the prediction of the feedback functions of facial expressions. This work builds upon and extends previous research on (a) the annotation and analysis of emotions in the corpus in which it was suggested that emotions are related to specific communicative functions, and (b) the prediction of feedback head movements using multimodal information. The results of the experiments show that information on multimodal behaviours which co-occur with the facial expressions improve the classifier performance. The improvement of the F-measure with respect to the unimodal baseline is of 0.269 and this result is parallel to that obtained for head movements in the same corpus. The experiments also show that the annotations of emotions contribute further to the prediction of feedback facial expressions confirming their relation. The best results are obtained training the classifier on the shape of facial expressions and co-occurring head movements, emotion labels, the gesturer’s and the interlocutor’s speech and can be used in applied systems.

Microblog text has always been the focus and difficulty of text emotion analysis due to its small size, strong timeliness and randomness. Traditional deep learning emotion analysis methods lack the focus on emotional knowledge such as emotional words, expressions, and degree adverbs in sentences. To this end, this paper proposes a dual-channel microblog emotion analysis method based on RoBERTa-WWM and multi-head attention. This paper use the extended emotional resource library to construct the emotional knowledge set of each sentence, and use the RoBERTa-WWM pre-training model suitable for Chinese text to obtain the feature representation of the sentence and the emotional knowledge set respectively. Then, the sentence feature representation and the emotional knowledge representation are inputted into a TextCNN-BiGRU network and a Multi-Head Attention network to obtain deeper semantics feature and attention feature of emotional knowledge. Finally, the semantic features and the emotional knowledge attention feature are combined to train the Microblog emotion analysis model. The experimental results on the NLPCC2014 data set show that it has achieved better results than the comparative experimental model. Experiments show that this method can effectively perform emotion analysis on Microblog texts.KeywordsDeep learningDual-channel text sentiment classificationRoBERTa-WWMMulti-Head AttentionBiGRU

Emotion Cause Pair Extraction By Multi Task Learning on Enhanced English Dataset

Facing fast-increasing electronic documents in the Digital Media Age, the need to extract textual features of online texts for better communication is growing. Sentiment classification might be the key method to catch emotions of online communication, and developing corpora with annotation of emotions is the first step to achieving sentiment classification. However, the labour-intensive and costly manual annotation has resulted in the lack of corpora for emotional words. Furthermore, single-label semantic corpora could hardly meet the requirement of modern analysis of complicated user’s emotions, but tagging emotional words with multiple labels is even more difficult than usual. Improvement of the methods of automatic emotion tagging with multiple emotion labels to construct new semantic corpora is urgently needed. Taking Twitter short texts as the case, this study proposes a new semi-automatic method to annotate Internet short texts with multiple labels and form a multi-labelled corpus for further algorithm training. Each sentence is tagged with both the emotional tendency and polarity, and each tweet, which generally contains several sentences, is tagged with the first two major emotional tendencies. The semi-automatic multi-labelled annotation is achieved through the process of selecting the base corpus and emotional tags, data preprocessing, automatic annotation through word matching and weight calculation, and manual correction in case of multiple emotional tendencies are found. The experiments on the Sentiment140 published Twitter corpus demonstrate the effectiveness of the proposed approach and show consistency between the results of semi-automatic annotation and manual annotation. By applying this method, this study summarises the annotation specification and constructs a multi-labelled emotion corpus with 6500 tweets for further algorithm training.