Natural Language Processing
As part of our ongoing work, we are involved in projects that are aimed at developing tools for multi-aspect text analysis in Slavic and Germanic languages. The NLP team creates tools that address universal observations about natural language on the one hand and take the specifics of each language into account on the other. To date, data sets that support language analysis and tools for text segmentation, the lemmatization of multi-segment expressions, keyword extraction, and text classification have been prepared.
The research includes the following:
- keyword search and organization based on text semantics;
- automatic text summarization;
- named entity recognition using knowledge-based approaches;
- development of expert systems;
- automatic question-answering.
Additionally, an international project called “example-based AI legal guidance (ExAILe)” is being conducted; this focuses on documents that are written in legalese and analyzes the meanings that are contained in these documents while developing dialog systems. The project aims to build an expert system that facilitates the interpretation of legal documents and is capable of providing substantive answers to questions that are posed in colloquial language.
keywords: natural language processing, lemmatization, text segmentation, NER (named entity recognition), expert systems, QA (question answering), dialog systems
Projects: https://lemkin.pl, https://arxiv.org/abs/1808.00926
Contact: dr inż. Aleksander Smywiński-Pohl BPP AGH
e-mail: apohllo@agh.edu.pl
Team:
dr Aleksander Smywiński-Pohl
mgr Magdalena Król
dr inż. Zbigniew Kaleta
inż. Paweł Lewkowicz
mgr inż. Damian Stachura
inż. Piotr Mirosław