Unifying Explicit and Implicit Feedback for Rating Prediction and Ranking Recommendation Tasks

This is the implementation of the following paper:

@InProceedings{unifying_exp0licit_implicit,
  author    = {Amir H. Jadidinejad and Craig Macdonald and Iadh Ounis},
  title     = {Unifying Explicit and Implicit Feedback for Rating Predictionand Ranking Recommendation Tasks},
  publisher = {ACM},
  booktitle = {International Conference on the Theory of Information Retrieval (ICTIR)},
  year      = {2019},
}

Requirements

• pytorch (1.0.1)
• spotlight (0.1.5)
• pytrec-eval (0.3)

Results

The following table summarizes the results of the proposed model ($\Phi_{D_{i}^{*}}$) compared to the baseline explicit ($\Phi_{D_{e}}$) and implicit ($MF_{bpr}$) models:

See the paper for more details.

How to reproduce?

Use the corresponding Jupyter notebook to reproduce the results of each dataset:

• GoodReads Children
• GoodReads Comics & Graphics
• GoodReads Fantasy & Paranormal
• Steam
• Douban
• Dianping