Dempster-Shafer Τheory Application in Recommender Systems and Comparison of Constraint Programming’s and Möbius Transform’s Implementation
In this repository you can find my Thesis for my Diploma (BSc) in Informatics and Telecommunications, from the National Kapodistrian University of Athens (https://pergamos.lib.uoa.gr/uoa/dl/frontend/el/browse/3228581). The title of the thesis is "Dempster-Shafer Τheory Application in Recommender Systems and Comparison of Constraint Programming’s and Möbius Transform’s Implementation".
The thesis with the corresponding technical APPENDIX (ANNEX in this case) is found in Thesis.pdf.
The implementation is in the folder implementation.
In order to execute the code ECLiPSe 6.1, Python 2.7 & 3.10, rpy2, PyCLP should be installed in the machine. If you find difficulties in installing the PyCLP please refer to the ANNEX in the Thesis-Report.
- The folder application_prolog refers to the Application of DST and use of CLP.
- The folder prolog_vs_ibelief refers to the comparison of the two implementations.
In each folder you will find a bash script (.sh) that runs multiple executions of the code.
For the code structure and the functionality of each module, please refer to the Thesis in Chapters 6 and 7 respectively.
The dataset won't be provided here, as it is large, but can be found in :
- https://drive.google.com/drive/folders/1tqFFT6QBJaE7qzGn44tnt4Yf5ygiB9ch?usp=sharing or
- https://www.kaggle.com/datasets/vodclickstream/netflix-audience-behaviour-uk-movies
The dataset should be placed inside application_prolog folder and saved with the name 'vodclickstream_uk_movies_03.csv' in order for the code to work.