A5010000036- Natural Language Processing Techniques
- Topic Modeling
- Anomaly Detection Techniques and Applications
- Speech and dialogue systems
- Semantic Web and Ontologies
- Machine Learning and Data Classification
- Advanced Text Analysis Techniques
- Autism Spectrum Disorder Research
- Data Stream Mining Techniques
- Human Pose and Action Recognition
- Innovation, Technology, and Society
- Software Engineering Research
- Language and cultural evolution
- Neural Networks and Applications
- Digital Economy and Work Transformation
- Privacy, Security, and Data Protection
- Digital and Cyber Forensics
- Context-Aware Activity Recognition Systems
- Blockchain Technology Applications and Security
- Religious Tourism and Spaces
- Logic, programming, and type systems
- Advanced Graph Neural Networks
- Algorithms and Data Compression
- Digitalization, Law, and Regulation
- Mobile Crowdsensing and Crowdsourcing
Otto Group (Germany)
2021
University of Roehampton
2020
University of Warwick
2020
Spiegel Institut (Germany)
2016
Saarland University
2008-2015
Recent work has shown that the integration of visual information into text-based models can substantially improve model predictions, but so far only extracted from static images been used. In this paper, we consider problem grounding sentences describing actions in videos. We present a general purpose corpus aligns high quality videos with multiple natural language descriptions portrayed videos, together an annotation how similar action are to each other. Experimental results demonstrate...
Seid Muhie Yimam, Heiner Ulrich, Tatiana von Landesberger, Marcel Rosenbach, Michaela Regneri, Alexander Panchenko, Franziska Lehmann, Uli Fahrer, Chris Biemann, Kathrin Ballweg. Proceedings of ACL-2016 System Demonstrations. 2016.
A broad-coverage corpus such as the Human Language Project envisioned by Abney and Bird (2010) would be a powerful resource for study of endangered languages. Existing corpora are limited in range languages covered, standardisation, or machine-readability. In this paper we present SeedLing, seed Project. We first survey existing efforts to compile cross-linguistic resources, then describe our own approach. To build foundation text Universal Corpus, crawl clean texts from several web sources...
Underspecification-based algorithms for processing partially disambiguated discourse structure must cope with extremely high numbers of readings. Based on previous work dominance graphs and weighted tree grammars, we provide the first possibility computing an underspecified description a best representation efficiently enough to process even longest discourses in RST Discourse Treebank.
Predictive geometric models deliver excellent results for many Machine Learning use cases. Despite their undoubted performance, neural predictive algorithms can show unexpected degrees of instability and variance, particularly when applied to large datasets. We present an approach measure changes in with respect both output consistency topological stability. Considering the example a recommender system using word2vec, we analyze influence single data points, approximation methods parameter...
We present a study about automated discourse analysis of oral narrative language in adolescents with autistic spectrum disorder (ASD).The basis this evaluation is an existing dataset fictional narrations individuals ASD and two matched comparison groups.We use three robust measures for quantifying different aspects text cohesion on corpus.These several combinations them correlate strongly human annotations.Our will show which these also distinguish the group from groups, do not, differences...
Different aspects of language processing have been shown to be sensitive priming but the findings studies examining effects in adolescents with Autism Spectrum Disorder (ASD) and Developmental Language (DLD) inconclusive. We present a study analysing visual implicit semantic ASD DLD. Based on dataset fictional script-like narratives, we evaluate how often extensively, content two different sources is used by participants. The first source was visual, consisting images participants assist...
This paper describes lex4all, an opensource PC application for the generation and evaluation of pronunciation lexicons in any language. With just a few minutes recorded audio no expert knowledge linguistics or speech technology, individuals organizations seeking to create speech-driven applications lowresource languages can build enabling recognition small vocabularies (up 100 terms, roughly) target language using existing engine designed high-resource source (e.g. English). To such...
We present an approach to compute the monetary value of individual data points, in context automated decision system. The proposed method enables us explore and implement a paradigm minimalism for large-scale machine learning systems. Data minimalistic implementations enhance scalability, while maintaining or even optimizing system's performance. Using two types recommender systems, we first demonstrate how much is ineffective both settings. then general account computing via sensitivity...