amirj / unifying_explicit_implicit

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

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