Reproducibility Study for CosRec: 2D Convolutional Neural Networks for Sequential Recommendation
P. Marko, Seljan Tomislav, Vrkic Ivan · Zenodo (CERN European Organization for Nuclear Research) · 2022
One of significant challenges in general research is to ensure that published scientific results are reliable and reproducible. Repro- ducibility (i.e. obtaining results as presented in a research using the same data and experiments) is a necessary step to verify the reliability of results. Reproducibility is also considered to be an crucial component in helping research as a whole, thereby allowing the scientists to quickly convert new findings to practice and progress with their work. The goal of this report is to verify the quantitative results and claims in the paper CosRec: 2D Convolutional Neural Networks for Sequential Recommendation by reproducing the described computational experiments and performing a detailed analysis.