Komputationelle Ereignisbewertung auf Basis von Appraisaltheorien für Emotionsanalyse (CEAT)

Emotionsanalyse wurde bisher in der Regel als Textklassifikationsaufgabe formuliert, in der vordefininierte Klassen Textsegmenten zugewiesen wurden. Die Klassen entsprechen typischerweise den Basisemotion (Wut, Angst, Freude, Überraschung, Traurigkeit, Ekel) oder (zusätzlich Vertrauen und Antizipation). Eine weitere Alternative stellt das Valenz-Arousal-Dominanz-Modell als Referenzsystem dar. Diese Ansätze stellen allerdings einen Unterschied in dem Stand der Forschung zwischen Psychologie und komputationeller Linguistik dar, da in dem erstgenannten Feld die Appraisaltheorien akzeptiert sind, aber bisher nie für Textanalyse genutzt wurden.Diesen Unterschied im Forschungsstand der verschiedenen Disziplinen verkleinern wir mit dem Projekt CEAT. Wir erstellen komputationelle Modelle auf Basis des kognitiven Appraisals von Ereignissen und, zu einem geringeren Maße, auf Beschreibungen von körperlichen Reaktionen und der Motivationskomponente von Emotionen. Als Basis für die Modellierung des kognitiven Appraisals nutzen wir die Arbeiten von Smith/Ellsworth (1985), welche zeigten, dass die Variablen wie angenehm ein Ereignis ist, wie verantwortlich man sich fühlt, wie sicher man ist, wieviel Aufmerksamkeit man dem Ereignis entgegenbringt und wieviel situationelle Kontrolle man hat, ausreichend sind um zwischen 15 Emotionen zu diskriminieren.In diesem Projekt erstellen wir zwei Modelle um diese Appraisaldimensionen textuellen Ereignisbeschreibungen zuzuweisen, zum einen auf Basis von semantischem Parsing, zum anderen auf Basis von tiefen neuronalen Netzen. Diese Dimensionen werden dann genutzt um die Emotion vorherzusagen, welche mit dem beschriebenen Ereignis wahrscheinlich verknüpft wird. Diese Modell werden erstmalig die Möglichkeit schaffen, Emotionen Ereignisbeschreibungen zuzuweisen, auch wenn Emotionsworte oder direkte Nennungen der Emotion nicht verfügbar sind.

CEAT ist ein Folgeprojekt zu SEAT (Strukturierte Emotionsanalyse in Text), in dem wir Emotionen als Ereignisse modelliert haben. Im Gegensatz dazu modellieren wir in CEAT Ereignisse und deren Wirkung auf Emotionen.

Das Projekt SEAT begann 2018. CEAT startete in 2021. Die Projekte werden von der Deutschen Forschungsgemeinschaft (DFG) finanziert.

Projekt-bezogene Publikationen

Wegge, Maximilian/Klinger, Roman (2024): Topic Bias in Emotion Classification. In: Proceedings of the Ninth Workshop on Noisy and User-generated Text (W-NUT 2024). San Ġiljan, Malta: Association for Computational Linguistics. S. 89–103.

Wemmer, Eileen/Labat, Sofie/Klinger, Roman (2024): EmoProgress: Cumulated Emotion Progression Analysis in Dreams and Customer Service Dialogues. In: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Torino, Italia: ELRA and ICCL. S. 5660–5677.

Klinger, Roman (2023a): Emotionsklassifikation in Texten unter Berücksichtigung des Komponentenprozessmodells. In: Jaki, Sylvia/Steiger, Stefan (Hg.): Digitale Hate Speech: interdisziplinäre Perspektiven auf Erkennung, Beschreibung und Regulation. Berlin, Heidelberg: J.B. Metzler. S. 131–154.

Klinger, Roman (2023b): Where are We in Event-centric Emotion Analysis?: Bridging Emotion Role Labeling and Appraisal-based Approaches. In: Proceedings of the Big Picture Workshop. Singapore: Association for Computational Linguistics. S. 1–17.

Troiano, Enrica/Klinger, Roman/Padó, Sebastian (2023): On the Relationship between Frames and Emotionality in Text. In: Northern European Journal of Language Technology 9.

Troiano, Enrica/Oberländer, Laura/Klinger, Roman (2023): Dimensional Modeling of Emotions in Text with Appraisal Theories: Corpus Creation, Annotation Reliability, and Prediction. In: Computational linguistics 49, S. 1–72.

Wegge, Maximilian/Klinger, Roman (2023): Automatic Emotion Experiencer Recognition. In: Proceedings of the 3rd Workshop on Computational Linguistics for the Political and Social Sciences. Ingolstadt: Association for Computational Lingustics. S. 1–7.

Štajner, Sanja/Klinger, Roman (2023): Emotion Analysis from Texts. In: Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: Tutorial Abstracts. Dubrovnik: Association for Computational Linguistics. S. 7–12.

Plaza-del-Arco, Flor Miriam/Martín-Valdivia, María-Teresa/Klinger, Roman (2022): Natural Language Inference Prompts for Zero-shot Emotion Classification in Text across Corpora. In: Proceedings of the 29th International Conference on Computational Linguistics. Gyeongju: International Committee on Computational Linguistics. S. 6805–6817.

Sabbatino, Valentino et al. (2022): “splink” is happy and “phrouth” is scary: Emotion Intensity Analysis for Nonsense Words. In: Proceedings of the 12th Workshop on Computational Approaches to Subjectivity, Sentiment & Social Media Analysis. Dublin: Association for Computational Linguistics. S. 37–50.

Troiano, Enrica et al. (2022): x-enVENT: A Corpus of Event Descriptions with Experiencer-specific Emotion and Appraisal Annotations. In: Proceedings of the Thirteenth Language Resources and Evaluation Conference. Marseille: European Language Resources Association. S. 1365–1375.

Wegge, Maximilian et al. (2022): Experiencer-Specific Emotion and Appraisal Prediction. In: Proceedings of the Fifth Workshop on Natural Language Processing and Computational Social Science (NLP+CSS). Abu Dhabi: Association for Computational Linguistics. S. 25–32.

Casel, Felix/Heindl, Amelie/Klinger, Roman (2021): Emotion Recognition under Consideration of the Emotion Component Process Model. In: Proceedings of the 17th Conference on Natural Language Processing (KONVENS 2021). KONVENS 2021 Organizers. S. 49–61.

Doan Dang, Bao Minh/Oberländer, Laura Ana Maria/Klinger, Roman (2021): Emotion Stimulus Detection in German News Headlines. In: Proceedings of the 17th Conference on Natural Language Processing (KONVENS 2021). KONVENS 2021 Organizers. S. 73–85.

Hofmann, Jan/Troiano, Enrica/Klinger, Roman (2021): Emotion-Aware, Emotion-Agnostic, or Automatic: Corpus Creation Strategies to Obtain Cognitive Event Appraisal Annotations. In: Proceedings of the Eleventh Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Association for Computational Linguistics. S. 160–170.

Troiano, Enrica/Padó, Sebastian/Klinger, Roman (2021): Emotion Ratings: How Intensity, Annotation Confidence and Agreements are Entangled. In: Proceedings of the Eleventh Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Association for Computational Linguistics. S. 40–49.

Bostan, Laura Ana Maria/Kim, Evgeny/Klinger, Roman (2020): GoodNewsEveryone: A Corpus of News Headlines Annotated with Emotions, Semantic Roles, and Reader Perception. In: LREC 2020 Marseille: Twelfth International Conference on Language Resources and Evaluation, May 11-16, 2020, Palais du Pharo, Marseille, France: conference proceedings. Paris: European Language Resources Association. S. 1554–1566.

Haider, Thomas et al. (2020): PO-EMO: Conceptualization, Annotation, and Modeling of Aesthetic Emotions in German and English Poetry. In: Proceedings of the Twelfth Language Resources and Evaluation Conference. S. 1652–1663.

Helbig, David/Troiano, Enrica/Klinger, Roman (2020): Challenges in Emotion Style Transfer: An Exploration with a Lexical Substitution Pipeline. In: Proceedings of the Eighth International Workshop on Natural Language Processing for Social Media. Association for Computational Linguistics. S. 41–50.

Hofmann, Jan et al. (2020): Appraisal Theories for Emotion Classification in Text. In: Proceedings of the 28th International Conference on Computational Linguistics. International Committee on Computational Linguistics. S. 125–138.

Klinger, Roman/Kim, Evgeny/Padó, Sebastian (2020): Emotion Analysis for Literary Studies. In: Reiter, Nils/Pichler, Axel/Kuhn, Jonas (Hg.): Reflektierte algorithmische Textanalyse: Interdisziplinäre(s) Arbeiten in der CRETA-Werkstatt. Berlin, Boston: De Gruyter. S. 237–268.

Oberländer, Laura Ana Maria/Klinger, Roman (2020): Token Sequence Labeling vs. Clause Classification for English Emotion Stimulus Detection. In: Proceedings of the Ninth Joint Conference on Lexical and Computational Semantics. Association for Computational Linguistics. S. 58–70.

Oberländer, Laura/Reich, Kevin/Klinger, Roman (2020): Experiencers, Stimuli, or Targets: Which Semantic Roles Enable Machine Learning to Infer the Emotions?. In: Proceedings of the Third Workshop on Computational Modeling of People’s Opinions, Personality, and Emotion’s in Social Media. Association for Computational Linguistics. S. 119–128.

Troiano, Enrica/Klinger, Roman/Padó, Sebastian (2020): Lost in Back-Translation: Emotion Preservation in Neural Machine Translation. In: Proceedings of the 28th International Conference on Computational Linguistics. International Committee on Computational Linguistics. S. 4340–4354.

Bostan, Laura/Klinger, Roman (2019): Exploring fine-tuned embeddings that model intensifiers for emotion analysis. In: Proceedings of the Tenth Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Minneapolis: Association for Computational Linguistics. S. 25–34.

Kim, Evgeny/Klinger, Roman (2019a): A Survey on Sentiment and Emotion Analysis for Computational Literary Studies. In: Zeitschrift fuer Digitale Geisteswissenschaften 4.

Kim, Evgeny/Klinger, Roman (2019b): Frowning Frodo, Wincing Leia, and a Seriously Great Friendship: Learning to Classify Emotional Relationships of Fictional Characters. In: Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. Association for Computational Linguistics. S. 647–653.

Kim, Evgeny/Klinger, Roman (2019c): An Analysis of Emotion Communication Channels in Fan-Fiction: Towards Emotional Storytelling. In: Proceedings of the Second Workshop on Storytelling. Association for Computational Linguistics. S. 56–64.

Troiano, Enrica/Padó, Sebastian/Klinger, Roman (2019): Crowdsourcing and Validating Event-focused Emotion Corpora for German and English. In: Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics. S. 4005–4011.

Bostan, Laura Ana Maria/Klinger, Roman (2018): An Analysis of Annotated Corpora for Emotion Classification in Text. In: Proceedings of the 27th International Conference on Computational Linguistics. Santa Fe, New Mexico: Association for Computational Linguistics. S. 2104–2119.

Kim, Evgeny/Klinger, Roman (2018): Who Feels What and Why?: Annotation of a Literature Corpus with Semantic Roles of Emotions. In: Proceedings of the 27th International Conference on Computational Linguistics. Association for Computational Linguistics. S. 1345–1359.

Klinger, Roman et al. (2018): IEST: WASSA-2018 Implicit Emotions Shared Task. In: Proceedings of the 9th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Brüssel: Association for Computational Linguistics. S. 31–42.

Strohm, Florian/Klinger, Roman (2018): An Empirical Analysis of the Role of Amplifiers, Downtoners, and Negations in Emotion Classification in Microblogs. In: 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA). Piscataway, NJ: IEEE.

Köper, Maximilian/Kim, Evgeny/Klinger, Roman (2017): IMS at EmoInt-2017: Emotion Intensity Prediction with Affective Norms, Automatically Extended Resources and Deep Learning. In: Proceedings of the 8th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Association for Computational Linguistics.

Schuff, Hendrik et al. (2017): Annotation, Modelling and Analysis of Fine-Grained Emotions on a Stance and Sentiment Detection Corpus. In: Proceedings of the 8th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis. Association for Computational Linguistics. S. 13–23.