Authors: Miha Bizjak, Anže Gregorc, Rok Grmek
This is a project for the Natural language processing course (FRI UL 2019/2020). Our task is to get the subjective information from text material that refer to a entity with the use of natural language processing and other methods. A entity is considered as a person, organization or a location and can be represented multiple times in one document or a sentence and there could be more entities in one document. All methods are evaluated on the SentiCoref 1.0 data set.
$ pip3 install -r requirements.txt
$ python3
>>> import nltk
>>> nltk.download('punkt')
Training of tested models takes a lot of time. Trained models are available for download here. Copy the downloaded models in the 'data/models' directory.
$ cd src
$ python3 main.py
All results produced by this source code are available in the 'results.txt' file.