Privacy Preserving Point-of-Interest Recommendation Using Decentralized Matrix Factorization
Authors: Chaochao Chen, Ziqi Liu, Peilin Zhao, Jun Zhou, Xiaolong Li
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on two real-world datasets demonstrate that, comparing with the classic and state-of-the-art latent factor models, DMF significantly improvements the recommendation performance in terms of precision and recall. |
| Researcher Affiliation | Industry | Chaochao Chen, Ziqi Liu, Peilin Zhao, Jun Zhou, Xiaolong Li AI Department, Ant Financial Services Group, China {chaochao.ccc, ziqiliu, peilin.zpl, jun.zhoujun, xl.li}@antfin.com |
| Pseudocode | Yes | Algorithm 1: Random Walk Enhanced Nearby Collaborative DMF Optimization |
| Open Source Code | No | The paper does not provide a link to source code or explicitly state that source code is made available. |
| Open Datasets | No | We use two real-world datasets, i.e., Foursquare and Alipay. Foursquare is a famous benchmark dataset for evaluating a POI recommendation model (Yang, Zhang, and Qu 2016). Our Alipay dataset is sampled from user-merchant offline check-in records during 2017/07/01 to 2017/07/31, and we also perform similar preprocess on it. The paper mentions these datasets but does not provide specific access links, DOIs, or repositories for the exact versions used in the experiments, nor formal citations with author/year for the Foursquare dataset in the context of access. |
| Dataset Splits | No | We randomly sample 90% as training set and the rest 10% as test set. The paper specifies a train/test split but does not explicitly mention a validation set or other specific splitting methodologies required for reproduction beyond this basic split. |
| Hardware Specification | No | The paper does not provide any specific hardware details (such as GPU/CPU models, processor types, or memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies or library versions used for its experiments. |
| Experiment Setup | Yes | During comparison, we set user regularizer α = 0.1, learning rate θ = 0.1, and the returned number of POI k {5, 10}. We also set the maximum number of neighbor N = 2, and the number of sampled unobserved ratings m = 3. After we build the user adjacent graph, we simply set wi,i = 1 to eliminate the effect of mapping function on model performance, since this is not the focus of this paper. For the latent factor dimension K, we vary its values in {5, 10, 15}. For the random walk distance D, we vary its values in {1, 2, 3, 4}. We also vary β and γ in {10 3, 10 2, 10 1, 100, 101} to study their effects on DMF. |