A5082450455- Topic Modeling
- Natural Language Processing Techniques
- Advanced Text Analysis Techniques
- Sentiment Analysis and Opinion Mining
- Semantic Web and Ontologies
- Speech and dialogue systems
- Mental Health via Writing
- Multimodal Machine Learning Applications
- Authorship Attribution and Profiling
- Emotion and Mood Recognition
- Video Analysis and Summarization
- Misinformation and Its Impacts
- Humor Studies and Applications
- Text Readability and Simplification
- Deception detection and forensic psychology
- Text and Document Classification Technologies
- Music and Audio Processing
- Digital Mental Health Interventions
- Human Pose and Action Recognition
- Web Data Mining and Analysis
- Diabetes Management and Research
- Complex Network Analysis Techniques
- Mental Health Research Topics
- Speech Recognition and Synthesis
- Diabetes Treatment and Management
University of Michigan–Ann Arbor
2016-2025
University of North Texas
2007-2024
Kaiser Permanente Center for Health Research
2024
Milken Institute
2024
George Washington University
2024
MedStar Health
2024
Kings County Hospital Center
2024
State University of New York
2024
Carolina Institute for NanoMedicine
2024
University of Colorado Denver
2024
This paper introduces the use of Wikipedia as a resource for automatic keyword extraction and word sense disambiguation, shows how this online encyclopedia can be used to achieve state-of-the-art results on both these tasks. The also two methods combined into system able automatically enrich text with links encyclopedic knowledge. Given an input document, identifies important concepts in corresponding pages. Evaluations show that annotations are reliable hardly distinguishable from manual...
Emotion recognition in conversations is a challenging task that has recently gained popularity due to its potential applications. Until now, however, large-scale multimodal multi-party emotional conversational database containing more than two speakers per dialogue was missing. Thus, we propose the Multimodal EmotionLines Dataset (MELD), an extension and enhancement of EmotionLines. MELD contains about 13,000 utterances from 1,433 dialogues TV-series Friends. Each utterance annotated with...
This paper describes experiments concerned with the automatic analysis of emotions in text. We describe construction a large data set annotated for six basic emotions: ANGER, DISGUST, FEAR, JOY, SADNESS and SURPRISE, we propose evaluate several knowledge-based corpusbased methods identification these
Emotion detection in conversations is a necessary step for number of applications, including opinion mining over chat history, social media threads, debates, argumentation mining, understanding consumer feedback live conversations, and so on. Currently systems do not treat the parties conversation individually by adapting to speaker each utterance. In this paper, we describe new method based on recurrent neural networks that keeps track individual party states throughout uses information...
The "Affective Text" task focuses on the classification of emotions and valence (positive/negative polarity) in news headlines, is meant as an exploration connection between lexical semantics. In this paper, we describe data set used evaluation results obtained by participating systems.
Eneko Agirre, Carmen Banea, Daniel Cer, Mona Diab, Aitor Gonzalez-Agirre, Rada Mihalcea, German Rigau, Janyce Wiebe. Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016). 2016.
Eneko Agirre, Carmen Banea, Claire Cardie, Daniel Cer, Mona Diab, Aitor Gonzalez-Agirre, Weiwei Guo, Rada Mihalcea, German Rigau, Janyce Wiebe. Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014). 2014.
Eneko Agirre, Carmen Banea, Claire Cardie, Daniel Cer, Mona Diab, Aitor Gonzalez-Agirre, Weiwei Guo, Iñigo Lopez-Gazpio, Montse Maritxalar, Rada Mihalcea, German Rigau, Larraitz Uria, Janyce Wiebe. Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015). 2015.
This paper presents an innovative unsupervised method for automatic sentence extraction using graph-based ranking algorithms. We evaluate the in context of a text summarization task, and show that results obtained compare favorably with previously published on established benchmarks.
This paper introduces a graph-based algorithm for sequence data labeling, using random walks on graphs encoding label dependencies. The is illustrated and tested in the context of an unsupervised word sense disambiguation problem, shown to significantly outperform accuracy achieved through individual assignment, as measured standard sense-annotated sets.
With more than 10,000 new videos posted online every day on social websites such as YouTube and Facebook, the internet is becoming an almost infinite source of information. One crucial challenge for coming decade to be able harvest relevant information from this constant flow multimodal data. This paper addresses task sentiment analysis, conducts proof-of-concept experiments that demonstrate a joint model integrates visual, audio, textual features can effectively used identify in Web videos....
The proliferation of misleading information in everyday access media outlets such as social feeds, news blogs, and online newspapers have made it challenging to identify trustworthy sources, thus increasing the need for computational tools able provide insights into reliability content. In this paper, we focus on automatic identification fake content news. Our contribution is twofold. First, introduce two novel datasets task detection, covering seven different domains. We describe...
Emotion is intrinsic to humans and consequently, emotion understanding a key part of human-like artificial intelligence (AI). recognition in conversation (ERC) becoming increasingly popular as new research frontier natural language processing (NLP) due its ability mine opinions from the plethora publicly available conversational data on platforms such Facebook, Youtube, Reddit, Twitter, others. Moreover, it has potential applications health-care systems (as tool for psychological analysis),...
Emotion recognition in conversations is crucial for building empathetic machines. Present works this domain do not explicitly consider the inter-personal influences that thrive emotional dynamics of dialogues. To end, we propose Interactive COnversational memory Network (ICON), a multimodal emotion detection framework extracts features from conversational videos and hierarchically models self- inter-speaker into global memories. Such memories generate contextual summaries which aid...
This paper presents a knowledge-based method for measuring the semantic-similarity of texts. While there is large body previous work focused on finding semantic similarity concepts and words, application these word-oriented methods to text has not been yet explored. In this paper, we introduce that combines word-to-word metrics into text-to-text metric, show outperforms traditional based lexical matching.
In this paper, we present initial experiments in the recognition of deceptive language. We introduce three data sets true and lying texts collected for purpose, show that automatic classification is a viable technique to distinguish between truth falsehood as expressed also method class-based feature analysis, which sheds some light on features are characteristic text.
Although research in other languages is increasing, much of the work subjectivity analysis has been applied to English data, mainly due large body electronic resources and tools that are available for this language. In paper, we propose evaluate methods can be employed transfer a repository across languages. Specifically, attempt leverage on and, by employing machine translation, generate Through comparative evaluations two different (Romanian Spanish), show automatic translation viable...
In this paper, we explore unsupervised techniques for the task of automatic short answer grading. We compare a number knowledge-based and corpus-based measures text similarity, evaluate effect domain size on measures, also introduce novel technique to improve performance system by integrating feedback from student answers. Overall, our significantly consistently outperforms other methods grading that have been proposed in past.
In this paper, we address the task of utterance level emotion recognition in conversations using commonsense knowledge. We propose COSMIC, a new framework that incorporates different elements such as mental states, events, and causal relations, build upon them to learn interactions between interlocutors participating conversation. Current state-of-theart methods often encounter difficulties context propagation, shift detection, differentiating related classes. By learning distinct...
Using multimodal sentiment analysis, the presented method integrates linguistic, audio, and visual features to identify in online videos. In particular, experiments focus on a new dataset consisting of Spanish videos collected from YouTube that are annotated for polarity.