A5026648166- Topic Modeling
- Natural Language Processing Techniques
- Semantic Web and Ontologies
- Misinformation and Its Impacts
- Biomedical Text Mining and Ontologies
- Stock Market Forecasting Methods
- Mobile Agent-Based Network Management
- Memory, Trauma, and Commemoration
- Advanced Text Analysis Techniques
- Imbalanced Data Classification Techniques
- Media Influence and Politics
- Text and Document Classification Technologies
- Data Mining Algorithms and Applications
- Spam and Phishing Detection
- Data Quality and Management
- Web Data Mining and Analysis
- Sentiment Analysis and Opinion Mining
- Data Stream Mining Techniques
Technical University of Gabrovo
2021-2023
Ontotext (Bulgaria)
2015-2023
University of Sheffield
2023
Bulgarian Academy of Sciences
2019-2021
Institute of Information and Communication Technologies
2019
Recently, much attention has been given to models for identifying rumors in social media. Features that are helpful automatic inference of credibility, veracity, reliability information have described. The ultimate goal is train classification able recognize future high-impact as early possible, before the event unfolds. generalization power greatly hindered by domain-dependent distributions features, an issue insufficiently discussed. Here we study a large dataset consisting rumor and...
This paper addresses the task of categorizing companies within industry classification schemes.The dataset consists encyclopedic articles about and their economic activities.The target schema is build by mapping linked open data in a semi-supervised manner.Target classes are built bottom-up from DBpedia.We apply several state art text techniques, based both on deep learning classical vectorspace models.
The last several years have seen a massive increase in the quantity and influence of disinformation being spread online.Various approaches been developed to target process at different stages from identifying sources tracking distribution social media providing follow up debunks people who encountered disinformation.One common conclusion each these is that too nuanced subjective topic for fully automated solutions work but data cross-reference high humans handle unassisted.Ultimately,...
This paper compares different solutions for the task of classifying companies with an industry classification scheme. Recent advances in deep learning methods show better performance text task. The dataset consists short textual descriptions and their economic activities. Target schemes are built by mapping related open data a semi-controlled manner. classes from bottom up DBpedia. For experiments used modifications BERT, XLNet, Glove ULMfit pre-trained models English. Two simple perceptron...
Despite rapid developments in the field of Natural Language Processing (NLP) past few years, task Multilingual Entity Linking (MEL) and especially its end-to-end formulation remains challenging.In this paper we aim to evaluate solutions for general end-toend multilingual entity linking by conducting experiments using both existing complete approaches novel combinations pipelines solving task.The results identify best performing current suggest some directions further research.