A5055258367- Natural Language Processing Techniques
- Speech and dialogue systems
- Topic Modeling
- Intelligent Tutoring Systems and Adaptive Learning
- Semantic Web and Ontologies
- Innovative Teaching and Learning Methods
- Multi-Agent Systems and Negotiation
- Advanced Text Analysis Techniques
- Online Learning and Analytics
- Social Robot Interaction and HRI
- Multimodal Machine Learning Applications
- Language, Metaphor, and Cognition
- AI-based Problem Solving and Planning
- Biomedical Text Mining and Ontologies
- Data Visualization and Analytics
- Syntax, Semantics, Linguistic Variation
- Teaching and Learning Programming
- Text Readability and Simplification
- Digital Mental Health Interventions
- Logic, Reasoning, and Knowledge
- Misinformation and Its Impacts
- Hate Speech and Cyberbullying Detection
- Music and Audio Processing
- Advanced Image and Video Retrieval Techniques
- Video Analysis and Summarization
University of Illinois Chicago
2015-2024
University of Illinois System
2022
University of Pittsburgh
1996-1998
Carnegie Mellon University
1994-1996
University of Pennsylvania
1990-1995
Philadelphia University
1993
University of Turin
1986-1987
In recent years, the kappa coefficient of agreement has become de facto standard for evaluating intercoder tagging tasks. this squib, we highlight issues that affect κ and community largely neglected. First, discuss assumptions underlying different computations expected component κ. Second, how prevalence bias measure.
Centering theory is the best-known framework for theorizing about local coherence and salience; however, its claims are articulated in terms of notions which only partially specified, such as “utterance,” “realization,” or “ranking.” A great deal research has attempted to arrive at more detailed specifications these parameters theory; a result, centering can be instantiated many different ways. We investigated systematic fashion effect on theory's ways setting parameters. Doing this...
This paper presents a first-order logic learning approach to determine rhetorical relations between discourse segments. Beyond linguistic cues and lexical information, our exploits compositional semantics segment structure data. We report statistically significant improvement in classifying over attribute-value paradigms such as Decision Trees, RIPPER Naive Bayes. For parsing, modified shift-reduce parsing model that uses relation classifier significantly outperforms right-branching...
Based on an ongoing attempt to integrate Natural Language instructions with human figure animation, we demonstrate that agents' understanding and use of can complement what they derive from the environment in which act. We focus two attitudes contribute behavior—their intentions their expectations—and shown how such ways environment. also show require more than one context interpretation thus evolve as activity progresses. A significant consequence is behavior cannot simply be treated “front...
In this paper we will present our ongoing work on a plan-based discourse processor developed in the context of Enthusiast Spanish to English translation system as part JANUS multi-lingual speech-to-speech system. We demonstrate that theories which postulate strict tree structure either intentional or attentional level are not totally adequate for handling spontaneous dialogues. extension approach along with its implementation processor. outperforms an based approach.
We discuss Feature Latent Semantic Analysis (FLSA), an extension to (LSA). LSA is a statistical method that ordinarily trained on words only; FLSA adds the richness of many other linguistic features corpus may be labeled with. applied dialogue act classification with excellent results. report results three corpora: CallHome Spanish, MapTask, and our own tutoring dialogues.
Physician and nurses have worked together for generations; however, their language training are vastly different; comparing contrasting work joint impact on patient outcomes is difficult in light of this difference. At the same time, EHR only includes physician perspective via physician-authored discharge summary, but not nurse documentation. Prior research area has focused collaboration usage similar terminology.
In this paper, I give an account, in terms of centering theory [GJW86], some phenomena pronominalization Italian, particular the use null or overt pronoun subject position. After a general introduction to Italian pronominal system, will review centering, and then show how original rules given [GJW86] have be extended modified. Finally, that does not account for two phenomena: first, functional role utterance may override predictions centering; second, can used refer whole discourse segment....
We developed two versions of a system, called iList, that helps students learn linked lists, an important topic in computer science curricula. The iList differ on the level feedback they can provide to students, specifically explanation syntax and execution errors. system has been fielded multiple classrooms institutions. Our results indicate is effective, considered interesting useful by its performance getting closer human tutors. Moreover, being context study tutoring, which guiding...
Natural Language Interfaces that facilitate data exploration tasks are rapidly gaining in interest the research community because they enable users to focus their attention on task of inquiry rather than mechanics chart construction. Yet, current systems rely solely processing user's explicit commands generate intended chart. These can be ambiguous due natural language tendencies such as speech disfluency and underspecification. In this paper, we developed studied how an always listening...
Our goal is to identify the features that predict occurrence and placement of discourse cues in tutorial explanations order aid automatic generation explanations. Previous attempts devise rules for text were based on intuition or small numbers constructed examples. We apply a machine learning program, C4.5, induce decision trees cue from corpus data coded variety previously thought affect usage. experiments enable us with most predictive power, show can be used useful generation.
Abhinav Kumar, Jillian Aurisano, Barbara Di Eugenio, Andrew Johnson, Alberto Gonzalez, Jason Leigh. Proceedings of the 17th Annual Meeting Special Interest Group on Discourse and Dialogue. 2016.
This paper presents an analysis of purpose clauses in the context instruction understanding. Such shows that goals affect interpretation and / or execution actions, lends support to proposal using generation enablement model relations between sheds light on some inference processes necessary interpret clauses.