Reproducible framework of Point-of-Interest recommendation systems on Docker
Description
This dataset provides all data and code required to reproduce the results achieved in the article "WERNECK, H. et al. Points of Interest recommendations: Methods, evaluation, and future directions. Information Systems". This dataset contains a Docker image (poi-rss.tar.gz) which allows the reproduction of [2]. The benchmark has several databases extracted from Yelp Dataset Challenge, evaluation metrics and recommendation systems. It also has the data extracted from a systematic literature review and distinct ways to visualize the data. The datasets used in the benchmark (benchmark-datasets.zip) and its raw and processed output files (benchmark-results.zip), in turn, are provided separately. The code is all under Creative Commons Attribution 4.0 International License and the benchmark datasets (benchmark-datasets.zip) are under Yelp terms of use, as described detailed in the PDF (Dataset_Challenge_Dataset_Agreement.pdf). Detailed instructions for this dataset are described in the article [1], which is under review for the time being. References: [1] Heitor Werneck, Nícollas Silva, Matheus Carvalho, Adriano C M Pereira, Fernando Mourão, and Leonardo Rocha. A Reproducible POI Recommendation Framework: Works Mapping and Benchmark Evaluation. 15. [2] Heitor Werneck, Nícollas Silva, Matheus Viana, Adriano C.M. Pereira, Fernando Mourão, and Leonardo Rocha. 2021. Points of Interest recommendations: Methods, evaluation, and future directions. Information Systems 101, (2021), 101789. DOI:https://doi.org/10/gmmq79
Files
Steps to reproduce
Since the artifact is a Docker image, any operating system that supports Docker will be able to reproduce the experiment. To reproduce our results, just follow the instructions below: (1) Download the poi-rss.tar.gz file. (2) Download and install Docker 20.10.7 or higher (possibly it also works with previous versions) (3) Load the image In the shell for example: $ docker load < poi-rss.tar.gz (4) Run and attach to the container In the shell for example: $ docker run -itd --name poi-rss poi-rss bash $ docker exec -it poi-rss bash (5) Run all experiments using the python scripts $ ./poi-rss-benchmark/run_all && ./poi-rss-map/run_all For more information, see article [1] mentioned above.