When humans judge the affective content of texts, they also implicitly assess the correctness of such judgment, that is, their confidence. We hypothesize that people’s (in)confidence that they performed well in an annotation task leads to (dis)agreements among each other. If this is true, confidence may serve as a diagnostic tool for systematic differences in annotations. To probe our assumption, we conduct a study on a subset of the Corpus of Contemporary American English, in which we ask raters to distinguish neutral sentences from emotion-bearing ones, while scoring the confidence of their answers. Confidence turns out to approximate inter-annotator disagreements. Further, we find that confidence is correlated to emotion intensity: perceiving stronger affect in text prompts annotators to more certain classification performances. This insight is relevant for modelling studies of intensity, as it opens the question wether automatic regressors or classifiers actually predict intensity, or rather human’s self-perceived confidence.

Appraisal theories explain how the cognitive evaluation of an event leads to a particular emotion. In contrast to theories of basic emotions or affect (valence/arousal), this theory has not received a lot of attention in natural language processing. Yet, in psychology it has been proven powerful: Smith and Ellsworth (1985) showed that the appraisal dimensions attention, certainty, anticipated effort, pleasantness, responsibility/control and situational control discriminate between (at least) 15 emotion classes. We study different annotation strategies for these dimensions, based on the event-focused enISEAR corpus (Troiano et al., 2019). We analyze two manual annotation settings: (1) showing the text to annotate while masking the experienced emotion label; (2) revealing the emotion associated with the text. Setting 2 enables the annotators to develop a more realistic intuition of the described event, while Setting 1 is a more standard annotation procedure, purely relying on text. We evaluate these strategies in two ways: by measuring inter-annotator agreement and by fine- tuning RoBERTa to predict appraisal variables. Our results show that knowledge of the emotion increases annotators’ reliability. Further, we evaluate a purely automatic rule-based labeling strategy (inferring appraisal from annotated emotion classes). Training on automatically assigned labels leads to a competitive performance of our classifier, even when tested on manual annotations. This is an indicator that it might be possible to automatically create appraisal corpora for every domain for which emotion corpora already exist.

The 2020 US Elections have been, more than ever before, characterized by social media campaigns and mutual accusations. We investigate in this paper if this manifests also in online communication of the supporters of the candidates Biden and Trump, by uttering hateful and offensive communication. We formulate an annotation task, in which we join the tasks of hateful/offensive speech detection and stance detection, and annotate 3000 Tweets from the campaign period, if they express a particular stance towards a candidate. Next to the established classes of favorable and against, we add mixed and neutral stances and also annotate if a candidate is mentioned with- out an opinion expression. Further, we an- notate if the tweet is written in an offensive style. This enables us to analyze if supporters of Joe Biden and the Democratic Party communicate differently than supporters of Donald Trump and the Republican Party. A BERT baseline classifier shows that the detection if somebody is a supporter of a candidate can be performed with high quality (.89 F1 for Trump and .91 F1 for Biden), while the detection that somebody expresses to be against a candidate is more challenging (.79 F1 and .64 F1, respectively). The automatic detection of hate/offensive speech remains challenging (with .53 F1). Our corpus is publicly available and constitutes a novel resource for computational modelling of offensive language under consideration of stances.

Social media contains unfiltered and unique information, which is potentially of great value, but, in the case of misinformation, can also do great harm. With regards to biomedical topics, false information can be particularly dangerous. Methods of automatic fact-checking and fake news detection address this problem, but have not been applied to the biomedical domain in social media yet. We aim to fill this research gap and annotate a corpus of 1200 tweets for implicit and explicit biomedical claims (the latter also with span annotations for the claim phrase). With this corpus, which we sample to be related to COVID-19, measles, cystic fibrosis, and depression, we develop baseline models which detect tweets that contain a claim automatically. Our analyses reveal that biomedical tweets are densely populated with claims (45 % in a corpus sampled to contain 1200 tweets focused on the domains mentioned above). Baseline classification experiments with embedding-based classifiers and BERT-based transfer learning demonstrate that the detection is challenging, however, shows acceptable performance for the identification of explicit expressions of claims. Implicit claim tweets are more challenging to detect.

Public companies are obliged to include financial and non-financial information within their cor- porate filings under Regulation S-K, in the United States (SEC, 2010). However, the requirements still allow for manager’s discretion. This raises the question to which extent the information is actually included and if this information is at all relevant for investors. We answer this question by training and evaluating an end-to-end deep learning approach (based on BERT and GloVe embeddings) to predict the financial and environmental performance of the company from the “Management’s Discussion and Analysis of Financial Conditions and Results of Operations” (MD&A) section of 10-K (yearly) and 10-Q (quarterly) filings. We further analyse the mediating effect of the environmental performance on the relationship between the company’s disclosures and financial performance. Hereby, we address the results of previous studies regarding environ- mental performance. We find that the textual information contained within the MD&A section does not allow for conclusions about the future (corporate) financial performance. However, there is evidence that the environmental performance can be extracted by natural language processing methods.

Emotion stimulus detection is the task of finding the cause of an emotion in a textual description, similar to target or aspect detection for sentiment analysis. Previous work approached this in three ways, namely (1) as text classification into an inventory of predefined possible stimuli (“Is the stimulus category A or B?”), (2) as sequence labeling of tokens (“Which tokens describe the stimulus?”), and (3) as clause classification (“Does this clause contain the emotion stimulus?”). So far, setting (3) has been evaluated broadly on Mandarin and (2) on English, but no comparison has been performed. Therefore, we analyze whether clause classification or token sequence labeling is better suited for emotion stimulus detection in English. We propose an integrated framework which enables us to evaluate the two different approaches comparably, implement models inspired by state-of-the-art approaches in Mandarin, and test them on four English data sets from different domains. Our results show that token sequence labeling is superior on three out of four datasets, in both clause-based and token sequence-based evaluation. The only case in which clause classification performs better is one data set with a high density of clause annotations. Our error analysis further confirms quantitatively and qualitatively that clauses are not the appropriate stimulus unit in English.

Span identification (in short, span ID) tasks such as chunking, NER, or code-switching detection, ask models to identify and classify relevant spans in a text. Despite being a staple of NLP, and sharing a common structure, there is little insight on how these tasks’ properties influence their difficulty, and thus little guidance on what model families work well on span ID tasks, and why. We analyze span ID tasks via performance prediction, estimating how well neural architectures do on different tasks. Our contributions are: (a) we identify key properties of span ID tasks that can inform performance prediction; (b) we carry out a large-scale experiment on English data, building a model to predict performance for unseen span ID tasks that can support architecture choices; (c), we investigate the parameters of the meta model, yielding new insights on how model and task properties interact to affect span ID performance. We find, e.g., that span frequency is especially important for LSTMs, and that CRFs help when spans are infrequent and boundaries non-distinctive.

We propose the task of emotion style transfer, which is particularly challenging, as emotions (here: anger, disgust, fear, joy, sadness, surprise) are on the fence between content and style. To understand the particular difficulties of this task, we design a transparent emotion style transfer pipeline based on three steps: (1) select the words that are promising to be substituted to change the emotion (with a brute-force approach and selection based on the attention mechanism of an emotion classifier), (2) find sets of words as candidates for substituting the words (based on lexical and distributional semantics), and (3) select the most promising combination of substitutions with an objective function which consists of components for content (based on BERT sentence embeddings), emotion (based on an emotion classifier), and fluency (based on a neural language model). This comparably straight-forward setup enables us to explore the task and understand in what cases lexical substitution can vary the emotional load of texts, how changes in content and style interact and if they are at odds. We further evaluate our pipeline quantitatively in an automated and an annotation study based on Tweets and find, indeed, that simultaneous adjustments of content and emotion are conflicting objectives: as we show in a qualitative analysis motivated by Scherer’s emotion component model, this is particularly the case for implicit emotion expressions based on cognitive appraisal or descriptions of bodily reactions.

Automatic emotion categorization has been predominantly formulated as text classification in which textual units are assigned to an emotion from a predefined inventory, for instance following the fundamental emotion classes proposed by Paul Ekman (fear, joy, anger, disgust, sadness, surprise) or Robert Plutchik (adding trust, anticipation). This approach ignores existing psychological theories to some degree, which provide explanations regarding the perception of events. For instance, the description that somebody discovers a snake is associated with fear, based on the appraisal as being an unpleasant and non-controllable situation. This emotion reconstruction is even possible without having access to explicit reports of a subjective feeling (for instance expressing this with the words “I am afraid.”). Automatic classification approaches therefore need to learn properties of events as latent variables (for instance that the uncertainty and the mental or physical effort associated with the encounter of a snake leads to fear). With this paper, we propose to make such interpretations of events explicit, following theories of cognitive appraisal of events, and show their potential for emotion classification when being encoded in classification models. Our results show that high quality appraisal dimension assignments in event descriptions lead to an improvement in the classification of discrete emotion categories. We make our corpus of appraisal-annotated emotion-associated event descriptions publicly available.

Machine translation provides powerful methods to convert text between languages, and is therefore a technology enabling a multilingual world. An important part of communication, however, takes place at the non-propositional level (e.g., politeness, formality, emotions), and it is far from clear whether current MT methods properly translate this information. This paper investigates the specific hypothesis that the non-propositional level of emotions is at least partially lost in MT. We carry out a number of experiments in a back-translation setup and establish that (1) emotions are indeed partially lost during translation; (2) this tendency can be reversed almost completely with a simple re-ranking approach informed by an emotion classifier, taking advantage of diversity in the n-best list; (3) the re-ranking approach can also be applied to change emotions, obtaining a model for emotion style transfer. An in-depth qualitative analysis reveals that there are recurring linguistic changes through which emotions are toned down or amplified, such as change of modality.

Emotion recognition is predominantly formulated as text classification in which textual units are assigned to an emotion from a predefined inventory (e.g., fear, joy, anger, disgust, sadness, surprise, trust, anticipation). More recently, semantic role labeling approaches have been developed to extract structures from the text to answer questions like: “who is described to feel the emotion?” (experiencer), “what causes this emotion?” (stimulus), and at which entity is it directed?” (target). Though it has been shown that jointly modeling stimulus and emotion category prediction is beneficial for both subtasks, it remains unclear which of these semantic roles enables a classifier to infer the emotion. Is it the experiencer, because the identity of a person is biased towards a particular emotion (X is always happy)? Is it a particular target (everybody loves X) or a stimulus (doing X makes everybody sad)? We answer these questions by training emotion classification models on five available datasets annotated with at least one semantic role by masking the fillers of these roles in the text in a controlled manner and find that across multiple corpora, stimuli and targets carry emotion information, while the experiencer might be considered a confounder. Further, we analyze if informing the model about the position of the role improves the classification decision. Particularly on literature corpora we find that the role information improves the emotion classification.

Obituaries contain information about people’s values across times and cultures, which makes them a useful resource for exploring cultural history. They are typically structured similarly, with sections corresponding to Personal Information, Biographical Sketch, Characteristics, Family, Gratitude, Tribute, Funeral Information and Other aspects of the person. To make this information available for further studies, we propose a statistical model which recognizes these sections. To achieve that, we collect a corpus of 20058 English obituaries from TheDaily Item, Remembering.CA and The London Free Press. The evaluation of our annotation guidelines with three annotators on 1008 obituaries shows a substantial agreement of Fleiss κ = 0.87. Formulated as an automatic segmentation task, a convolutional neural network outperforms bag-of-words and embedding-based BiLSTMs and BiLSTM-CRFs with a micro F1 = 0.81.

Most research on emotion analysis from text focuses on the task of emotion classification or emotion intensity regression. Fewer works address emotions as a phenomenon to be tackled with structured learning, which can be explained by the lack of relevant datasets. We fill this gap by releasing a dataset of 5000 English news headlines annotated via crowdsourcing with their associated emotions, the corresponding emotion experiencers and textual cues, related emotion causes and targets, as well as the reader’s perception of the emotion of the headline. This annotation task is comparably challenging, given the large number of classes and roles to be identified. We therefore propose a multiphase annotation procedure in which we first find relevant instances with emotional content and then annotate the more fine-grained aspects. Finally, we develop a baseline for the task of automatic prediction of semantic role structures and discuss the results. The corpus we release enables further research on emotion classification, emotion intensity prediction, emotion cause detection, and supports further qualitative studies.

Most approaches to emotion analysis of social media, literature, news, and other domains focus exclusively on basic emotion categories as defined by Ekman or Plutchik. However, art (such as literature) enables engagement in a broader range of more complex and subtle emotions. These have been shown to also include mixed emotional responses. We consider emotions in poetry as they are elicited in the reader, rather than what is expressed in the text or intended by the author. Thus, we conceptualize a set of aesthetic emotions that are predictive of aesthetic appreciation in the reader, and allow the annotation of multiple labels per line to capture mixed emotions within their context. We evaluate this novel setting in an annotation experiment both with carefully trained experts and via crowdsourcing. Our annotation with experts leads to an acceptable agreement of k = .70, resulting in a consistent dataset for future large scale analysis. Finally, we conduct first emotion classification experiments based on BERT, showing that identifying aesthetic emotions is challenging in our data, with up to .52 F1-micro on the German subset. Data and resources are available at https://github.com/tnhaider/poetry-emotion.

Sentiment analysis has a range of corpora available across multiple languages. For emotion analysis, the situation is more limited, which hinders potential research on crosslingual modeling and the development of predictive models for other languages. In this paper, we fill this gap for German by constructing deISEAR, a corpus designed in analogy to the well-established English ISEAR emotion dataset. Motivated by Scherer’s appraisal theory, we implement a crowdsourcing experiment which consists of two steps. In step 1, participants create descriptions of emotional events for a given emotion. In step 2, five annotators assess the emotion expressed by the texts. We show that transferring an emotion classification model from the original English ISEAR to the German crowdsourced deISEAR via machine translation does not, on average, cause a performance drop.

Adjective phrases like “a little bit surprised”, “completely shocked”, or “not stunned at all” are not handled properly by current state-of-the-art emotion classification and intensity prediction systems. Based on this finding, we analyze differences between embeddings used by these systems in regard to their capability of handling such cases and argue that intensifiers in context of emotion words need special treatment, as is established for sentiment polarity classification, but not for more fine-grained emotion prediction. To resolve this issue, we analyze different aspects of a post-processing pipeline which enriches the word representations of such phrases. This includes expansion of semantic spaces at the phrase level and sub-word level followed by retrofitting to emotion lexicons. We evaluate the impact of these steps with ‘A La Carte and Bag-of-Substrings extensions based on pretrained GloVe,Word2vec, and fastText embeddings against a crowd-sourced corpus of intensity annotations for tweets containing our focus phrases. We show that the fastText-based models do not gain from handling these specific phrases under inspection. For Word2vec embeddings, we show that our post-processing pipeline improves the results by up to 8% on a novel dataset densly populated with intensifiers while it does not decrease the performance on the established EmoInt dataset.

Centrality of emotion for the stories told by humans is underpinned by numerous studies in literature and psychology. The research in automatic storytelling has recently turned towards emotional storytelling, in which characters’ emotions play an important role in the plot development (Theune et al., 2004; y Perez, 2007; Mendez et al., 2016). However, these studies mainly use emotion to generate propositional statements in the form “A feels affection towards B” or “A confronts B”. At the same time, emotional behavior does not boil down to such propositional descriptions, as humans display complex and highly variable patterns in communicating their emotions, both verbally and non-verbally. In this paper, we analyze how emotions are expressed non-verbally in a corpus of fan fiction short stories. Our analysis shows that stories written by humans convey character emotions along various non-verbal channels. We find that some non-verbal channels, such as facial expressions and voice characteristics of the characters, are more strongly associated with joy, while gestures and body postures are more likely to occur with trust. Based on our analysis, we argue that automatic storytelling systems should take variability of emotion into account when generating descriptions of characters’ emotions.

The development of a fictional plot is centered around characters who closely interact with each other forming dynamic social networks. In literature analysis, such networks have mostly been analyzed without particular relation types or focusing on roles which the characters take with respect to each other. We argue that an important aspect for the analysis of stories and their development is the emotion between characters. In this paper, we combine these aspects into a unified framework to classify emotional relationships of fictional characters. We formalize it as a new task and describe the annotation of a corpus, based on fan-fiction short stories. The extraction pipeline which we propose consists of character identification (which we treat as given by an oracle here) and the relation classification. For the latter, we provide results using several approaches previously proposed for relation identification with neural methods. The best result of 0.45 F1 is achieved with a GRU with character position indicators on the task of predicting undirected emotion relations in the associated social network graph.

The automatic detection of satire vs. regular news is relevant for downstream applications (for instance, knowledge base population) and to improve the understanding of linguistic characteristics of satire. Recent approaches build upon corpora which have been labeled automatically based on article sources. We hypothesize that this encourages the models to learn characteristics for different publication sources (e.g., “The Onion” vs. “The Guardian”) rather than characteristics of satire, leading to poor generalization performance to unseen publication sources. We therefore propose a novel model for satire detection with an adversarial component to control for the confounding variable of publication source. On a large novel data set collected from German news (which we make available to the research community), we observe comparable satire classification performance and, as desired, a considerable drop in publication classification performance with adversarial training. Our analysis shows that the adversarial component is crucial for the model to learn to pay attention to linguistic properties of satire.

Most machine learning systems for natural language processing are tailored to specific tasks. As a result, comparability of models across tasks is missing and their applicability to new tasks is limited. This affects end users without machine learning experience as well as model developers. To address these limitations, we present DERE, a novel framework for declarative specification and compilation of template-based information extraction. It uses a generic specification language for the task and for data annotations in terms of spans and frames. This formalism enables the representation of a large variety of natural language processing challenges. The backend can be instantiated by different models, following different paradigms. The clear separation of frame specification and model backend will ease the implementation of new models and the evaluation of different models across different tasks. Furthermore, it simplifies transfer learning, joint learning across tasks and/or domains as well as the assessment of model generalizability. DERE is available as open-source software.

Sentiment analysis in low-resource languages suffers from a lack of annotated corpora to estimate high-performing models. Machine translation and bilingual word embeddings provide some relief through cross-lingual sentiment approaches. However, they either require large amounts of parallel data or do not sufficiently capture sentiment information. We introduce Bilingual Sentiment Embeddings (BLSE), which jointly represent sentiment information in a source and target language. This model only requires a small bilingual lexicon, a source-language corpus annotated for sentiment, and monolingual word embeddings for each language. We perform experiments on three language combinations (Spanish, Catalan, Basque) for sentence-level cross-lingual sentiment classification and find that our model significantly outperforms state-of-the-art methods on four out of six experimental setups, as well as capturing complementary information to machine translation. Our analysis of the resulting embedding space provides evidence that it represents sentiment information in the resource-poor target language without any annotated data in that language.

Supervised machine learning algorithms require training data whose generation for complex relation extraction tasks tends to be difficult. Being optimized for relation extraction at sentence level, many annotation tools lack in facilitating the annotation of relational structures that are widely spread across the text. This leads to non-intuitive and cumbersome visualizations, making the annotation process unnecessarily time-consuming. We propose SANTO, an easy-to-use, domain-adaptive annotation tool specialized for complex slot filling tasks which may involve problems of cardinality and referential grounding. The web-based architecture enables fast and clearly structured annotation for multiple users in parallel. Relational structures are formulated as templates following the conceptualization of an underlying ontology. Further, import and export procedures of standard formats enable interoperability with external sources and tools.

Domain adaptation for sentiment analysis is challenging due to the fact that supervised classifiers are very sensitive to changes in domain. The two most prominent approaches to this problem are structural correspondence learning and autoencoders. However, they either require long training times or suffer greatly on highly divergent domains. Inspired by recent advances in cross-lingual sentiment analysis, we provide a novel perspective and cast the domain adaptation problem as an embedding projection task. Our model takes as input two mono-domain embedding spaces and learns to project them to a bi-domain space, which is jointly optimized to (1) project across domains and to (2) predict sentiment. We perform domain adaptation experiments on 20 source-target domain pairs for sentiment classification and report novel state-of-the-art results on 11 domain pairs, including the Amazon domain adaptation datasets and SemEval 2013 and 2016 datasets. Our analysis shows that our model performs comparably to state-of-the-art approaches on domains that are similar, while performing significantly better on highly divergent domains. Our code is available at https://github.com/jbarnesspain/domain_blse

Most approaches to emotion analysis in fictional texts focus on detecting the emotion expressed in text. We argue that this is a simplification which leads to an overgeneralized interpretation of the results, as it does not take into account who experiences an emotion and why. Emotions play a crucial role in the interaction between characters and the events they are involved in. Until today, no specific corpora that capture such an interaction were available for literature. We aim at filling this gap and present a publicly available corpus based on Project Gutenberg, REMAN (Relational EMotion ANnotation), manually annotated for spans which correspond to emotion trigger phrases and entities/events in the roles of experiencers, targets, and causes of the emotion. We provide baseline results for the automatic prediction of these relational structures and show that emotion lexicons are not able to encompass the high variability of emotion expressions and demonstrate that statistical models benefit from joint modeling of emotions with its roles in all subtasks. The corpus that we provide enables future research on the recognition of emotions and associated entities in text. It supports qualitative literary studies and digital humanities. The corpus is available at http://www.ims.uni-stuttgart.de/data/reman .

Several datasets have been annotated and published for classification of emotions. They differ in several ways: (1) the use of different annotation schemata (e. g., discrete label sets, including joy, anger, fear, or sadness or continuous values including valence, or arousal), (2) the domain, and, (3) the file formats. This leads to several research gaps: supervised models often only use a limited set of available resources. Additionally, no previous work has compared emotion corpora in a systematic manner. We aim at contributing to this situation with a survey of the datasets, and aggregate them in a common file format with a common annotation schema. Based on this aggregation, we perform the first cross-corpus classification experiments in the spirit of future research enabled by this paper, in order to gain insight and a better understanding of differences of models inferred from the data. This work also simplifies the choice of the most appropriate resources for developing a model for a novel domain. One result from our analysis is that a subset of corpora is better classified with models trained on a different corpus. For none of the corpora, training on all data altogether is better than using a subselection of the resources. Our unified corpus is available at http://www.ims.uni-stuttgart.de/data/unifyemotion.

Past shared tasks on emotions use data with both overt expressions of emotions (I am so happy to see you!) as well as subtle expressions where the emotions have to be inferred, for instance from event descriptions. Further, most datasets do not focus on the cause or the stimulus of the emotion. Here, for the first time, we propose a shared task where systems have to predict the emotions in a large automatically labeled dataset of tweets without access to words denoting emotions. Based on this intention, we call this the Implicit Emotion Shared Task (IEST) because the systems have to infer the emotion mostly from the context. Every tweet has an occurrence of an explicit emotion word that is masked. The tweets are collected in a manner such that they are likely to include a description of the cause of the emotion – the stimulus. Altogether, 30 teams submitted results which range from macro F1 scores of 21 % to 71 %. The baseline (Max-Ent bag of words and bigrams) obtains an F1 score of 60 % which was available to the participants during the development phase. A study with human annotators suggests that automatic methods outperform human predictions, possibly by honing into subtle textual clues not used by humans. Corpora, resources, and results are available at the shared task website at http://implicitemotions.wassa2018.com.

Literary genres are commonly viewed as being defined in terms of content and stylistic features. In this paper, we focus on one particular class of lexical features, namely emotion information, and investigate the hypothesis that emotion-related information correlates with particular genres. Using genre classification as a testbed, we compare a model that computes lexicon-based emotion scores globally for complete stories with a model that tracks emotion arcs through stories on a subset of Project Gutenberg with five genres. Our main findings are: (a), the global emotion model is competitive with a large-vocabulary bag-of-words genre classifier (80%F1); (b), the emotion arc model shows a lower performance (59 % F1) but shows complementary behavior to the global model, as indicated by a very good performance of an oracle model (94 % F1) and an improved performance of an ensemble model (84 % F1); (c), genres differ in the extent to which stories follow the same emotional arcs, with particularly uniform behavior for anger (mystery) and fear (adventures, romance, humor, science fiction).

There has been a good amount of progress in sentiment analysis over the past 10 years, including the proposal of new methods and the creation of benchmark datasets. In some papers, however, there is a tendency to compare models only on one or two datasets, either because of time restraints or because the model is tailored to a specific task. Accordingly, it is hard to understand how well a certain model generalizes across different tasks and datasets. In this paper, we contribute to this situation by comparing several models on six different benchmarks, which belong to different domains and additionally have different levels of granularity (binary, 3-class, 4-class and 5-class). We show that Bi-LSTMs perform well across datasets and that both LSTMs and Bi-LSTMs are particularly good at fine-grained sentiment tasks (i.e., with more than two classes). Incorporating sentiment information into word embeddings during training gives good results for datasets that are lexically similar to the training data. With our experiments, we contribute to a better understanding of the performance of different model architectures on different data sets. Consequently, we detect novel state-of-the-art results on the SenTube datasets.

There is a rich variety of data sets for sentiment analysis (viz., polarity and subjectivity classification). For the more challenging task of detecting discrete emotions following the definitions of Ekman and Plutchik, however, there are much fewer data sets, and notably no resources for the social media domain. This paper contributes to closing this gap by extending the SemEval 2016 stance and sentiment datasetwith emotion annotation. We (a) analyse annotation reliability and annotation merging; (b) investigate the relation between emotion annotation and the other annotation layers (stance, sentiment); (c) report modelling results as a baseline for future work.

Social media are used by an increasing number of political actors. A small subset of these is interested in pursuing extremist motives such as mobilization, recruiting or radicalization activities. In order to counteract these trends, online providers and state institutions reinforce their monitoring efforts, mostly relying on manual workflows. We propose a machine learning approach to support manual attempts towards identifying right-wing extremist content in German Twitter profiles. Based on a fine-grained conceptualization of right-wing extremism, we frame the task as ranking each individual profile on a continuum spanning different degrees of right-wing extremism, based on a nearest neighbour approach. A quantitative evaluation reveals that our ranking model yields robust performance (up to 0.81 F₁ score) when being used for predicting discrete class labels. At the same time, the model provides plausible continuous ranking scores for a small sample of borderline cases at the division of right-wing extremism and New Right political movements.

Our submission to the WASSA-2017 shared task on the prediction of emotion intensity in tweets is a supervised learning method with extended lexicons of affective norms. We combine three main information sources in a random forrest regressor, namely (1), manually created resources, (2) automatically extended lexicons, and (3) the output of a neural network (CNN-LSTM) for sentence regression. All three feature sets perform similarly well in isolation (≈ .67 macro average Pearson correlation). The combination achieves .72 on the official test set (ranked 2nd out of 22 participants). Our analysis reveals that performance is increased by providing cross-emotional intensity predictions. The automatic extension of lexicon features benefit from domain specific embeddings. Complementary ratings for affective norms increase the impact of lexicon features. Our resources (ratings for 1.6 million twitter specific words) and our implementation is publicly available at http://www.ims.uni-stuttgart.de/data/ims_emoint.

Systems which build on top of information extraction are typically challenged to extract knowledge that, while correct, is not yet well-known. We hypothesize that a good confidence measure for relational information has the property that such interesting information is found between information extracted with very high confidence and very low confidence. We discuss confidence estimation for the domain of biomedical protein-protein relation discovery in biomedical literature. As facts reported in papers take some time to be validated and recorded in biomedical databases, such task gives rise to large quantities of unknown but potentially true candidate relations. It is thus important to rank them based on supporting evidence rather than discard them. In this paper, we discuss this task and propose different approaches for confidence estimation and a pipeline to evaluate such methods. We show that the most straight-forward approach, a combination of different confidence measures from pipeline modules seems not to work well. We discuss this negative result and pinpoint potential future research directions.

The automatic analysis of texts containing opinions of users about, e.g., products or political views has gained attention within the last decades. However, previous work on the task of analyzing user reviews about mobile applications in app stores is limited. Publicly available corpora do not exist, such that a comparison of different methods and models is difficult. We fill this gap by contributing the Sentiment Corpus of App Reviews (SCARE), which contains fine-grained annotations of application aspects, subjective (evaluative) phrases and relations between both. This corpus consists of 1,760 annotated application reviews from the Google Play Store with 2,487 aspects and 3,959 subjective phrases. We describe the process and methodology how the corpus was created. The Fleiss Kappa between four annotators reveals an agreement of 0.72. We provide a strong baseline with a linear-chain conditional random field and word-embedding features with a performance of 0.62 for aspect detection and 0.63 for the extraction of subjective phrases. The corpus is available to the research community to support the development of sentiment analysis methods on mobile application reviews.

Opinion mining has received wide attention in recent years. Models for this task are typically trained or evaluated with a manually annotated dataset. However, fine-grained annotation of sentiments including information about aspects and their evaluation is very labour-intensive. The data available so far is limited. Contributing to this situation, this paper describes the Bielefeld University Sentiment Analysis Corpus for German and English (USAGE), which we offer freely to the community and which contains the annotation of product reviews from Amazon with both aspects and subjective phrases. It provides information on segments in the text which denote an aspect or a subjective evaluative phrase which refers to the aspect. Relations and coreferences are explicitly annotated. This dataset contains 622 English and 611 German reviews, allowing to investigate how to port sentiment analysis systems across languages and domains. We describe the methodology how the corpus was created and provide statistics including inter-annotator agreement. We further provide figures for a baseline system and results for German and English as well as in a cross-domain setting. The results are encouraging in that they show that aspects and phrases can be extracted robustly without the need of tuning to a particular type of products.