A5088044797- Natural Language Processing Techniques
- Topic Modeling
- Semantic Web and Ontologies
- Web Data Mining and Analysis
- Advanced Text Analysis Techniques
- Software Engineering Research
- Biomedical Text Mining and Ontologies
- Speech and dialogue systems
- Text and Document Classification Technologies
- Service-Oriented Architecture and Web Services
- Advanced Database Systems and Queries
- Sentiment Analysis and Opinion Mining
- Algorithms and Data Compression
- Authorship Attribution and Profiling
- Cultural Differences and Values
- Cultural Heritage Management and Preservation
- Marine and fisheries research
- Ichthyology and Marine Biology
- Machine Learning and Algorithms
- Museums and Cultural Heritage
- Artificial Intelligence in Healthcare and Education
- Spam and Phishing Detection
- Data Stream Mining Techniques
- Conservation Techniques and Studies
- Explainable Artificial Intelligence (XAI)
National Centre of Scientific Research "Demokritos"
2015-2024
Institute of Informatics & Telecommunications
2011-2024
Aristotle University of Thessaloniki
2023
Hippocration General Hospital
2023
Institute of Informatics of the Slovak Academy of Sciences
2004-2019
Technical University of Darmstadt
2019
East Stroudsburg University
2019
University of Illinois Urbana-Champaign
2019
Microsoft (United States)
2019
Paris-Est Sup
2019
Argument extraction is the task of identifying arguments, along with their components in text. Arguments can be usually decomposed into a claim and one or more premises justifying it. The proposed approach tries to identify segments that represent argument elements (claims premises) on social Web texts (mainly news blogs) Greek language, for small set thematic domains, including articles politics, economics, culture, various issues, sports. exploits distributed representations words,...
This paper presents a method that assists in maintaining rule-based named-entity recognition and classification system. The underlying idea is to use separate system, constructed with the of machine learning, monitor performance training data for second system generated thus avoiding need manual tagging. disagreement two systems acts as signal updating generality approach illustrated by applying it large corpora different languages: Greek French. results are very encouraging, showing this...
In this paper we examine the application of an unsupervised extractive summarisation algorithm, TextRank, on a different task, identification argumentative components.Our main motivation is to whether there any potential overlap between and argument mining, approaches used in (which typically model document as whole) can have positive effect tasks mining.Evaluation has been performed two corpora containing user posts from on-line debating forum persuasive essays.Evaluation results suggest...
The recent success of distributed and dynamic infrastructures for knowledge sharing has raised the need semiautomatic/automatic ontology evolution strategies. Ontology is generally defined as timely adaptation an to changing requirements consistent propagation changes dependent artifacts. In this article, we present approach in context multimedia interpretation. relies on results obtained through reasoning interpretation resources, population with new individuals or enrichment concepts...
The recognition of Proper Nouns (PNs) is considered an important task in the area Information Retrieval and Extraction. However high performance most existing PN classifiers heavily depends upon availability large dictionaries domain-specific Nouns, a certain amount manual work for rule writing or tagging. Though it not heavy requirement to rely on some dictionary (often these resources are available web), its coverage domain corpus may be rather low, absence updating. In this paper we...
Argument extraction is the task of identifying arguments, along with their components in text. Arguments can be usually decomposed into a claim and one or more premises justifying it. Among novel aspects this work thematic domain itself which relates to Social Media, contrast traditional research area, concentrates mainly on law documents scientific publications. The huge increase social media communities, user tendency debate, makes identification arguments these texts necessity. from Media...
Every year, marine scientists around the world read thousands of otolith or scale images to determine age structure commercial fish stocks. This knowledge is important for fisheries and conservation management. However, age-reading procedure time-consuming costly perform due specialized expertise labor needed identify annual growth zones in otoliths. Effective automated systems are increase throughput reduce cost. DeepOtolith an open-source artificial intelligence (AI) platform that...
The field of exploration the values embraced by people and societies seems to be vast, as there are many different parameters examination. importance in formation human character and, extension, better is a key thematic area that date, has attracted lot research interest across various scientific domains. However, one wonders whether differ at an individual societal level; if they transform what extent this transformation impacts (and/or impacted) differently level. Although existing...
Handling missing values in a dataset is long-standing issue across many disciplines. Missing can arise from different sources such as mishandling of samples, measurement errors, lack responses, or deleted values. The main problem emerging this situation that algorithms can't run with incomplete datasets. Several methods exist for handling values, including "SoftImpute", "k-nearest neighbor", "mice", "MatrixFactorization", and "miss- Forest". However, performance comparisons these are hard to...
This paper presents our participation to the "Human Value Detection shared task (Kiesel et al., 2023), as "Andronicus of Rhodes. We describe approaches behind each entry in official evaluation, along with motivation approach. Our best-performing approach has been based on BERT large, 4 classification heads, implementing two different (with activation and loss functions), partitioning training data, handle class imbalance. Classification is performed through majority voting. The proposed...