Modeling Dynamic Missingness of Implicit Feedback for Recommendation
Authors: Menghan Wang, Mingming Gong, Xiaolin Zheng, Kun Zhang
NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on real-world datasets demonstrate the superiority of our method against state-of-the-art recommender systems. |
| Researcher Affiliation | Academia | Menghan Wang College of Computer Science, Zhejiang University wangmengh@zju.edu.cn Mingming Gong Department of Biomedical Informatics, University of Pittsburgh mig73@pitt.edu Xiaolin Zheng College of Computer Science, Zhejiang University xlzheng@zju.edu.cn Kun Zhang Department of Philosophy, Carnegie Mellon University kunz1@cmu.edu |
| Pseudocode | No | The paper describes the model and inference steps mathematically but does not include structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. |
| Open Datasets | Yes | We evaluate the performance of our method on three real-world datasets: 1) Movie Lens-100K dataset... 2) Movie Lens-1M dataset... 3) Last FM dataset... We transform the two Movie Lens datasets into implicit data by setting ratings that are >=3 to 1 and the others to 0. |
| Dataset Splits | Yes | We split the dataset for experiments with the following strategy: we first sort the historical ratings of each user by time order. Then the last records of users are used as test data, the second last records are used as validation data, and the remaining records are used for training. |
| Hardware Specification | Yes | The experiments are performed on a workstation with Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz, 256GB memory and two NVIDIA Pascal P100 GPUs. |
| Software Dependencies | No | All algorithms are implemented with Python and PyTorch. The paper does not specify version numbers for Python or PyTorch. |
| Experiment Setup | Yes | For PMF, we set K = 10. For WMF, we set K = 10, α = 0.4. For Expo MF, we set λθ = 0.01, λβ = 0.01, λy = 0.01, and K = 30. For our models, we set λθ = 0.1, λβ = 0.1, λy = 0.1, K = 30, ad ini = 1, and bd ini = 2. For item constraints, λOuter is set as 1, and λInner is set 10, 1, and 0.1 for Movie Lens100K, Movie Lens-1M, and Last FM, respectively. |