Personalized Time-Aware Tag Recommendation

Authors: Keqiang Wang, Yuanyuan Jin, Haofen Wang, Hongwei Peng, Xiaoling Wang

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental results show that our proposed model outperforms the state of the art tag recommendation methods in accuracy and has better ability to recommend new tags.
Researcher Affiliation Collaboration 1International Research Center of Trustworthy Software Shanghai Key Laboratory of Trustworthy Computing East China Normal University, Shanghai, China 2 Shenzhen Gowild Robotics Co. Ltd
Pseudocode Yes Algorithm 1: An Optimization Algorithm for TAPITF.
Open Source Code No The paper does not explicitly state that the source code for their methodology is released, nor does it provide a link to a code repository.
Open Datasets Yes We evaluate the models on the three public data sets Movielens, Last FM and Delicious described in table 1.
Dataset Splits No We use leave-one-out to split data set into train set and test set, which is that for each user, his tagging records on a certain item are randomly removed from the training set Strain and put into the test set St.
Hardware Specification No The paper discusses 'running time' (Table 3) but does not provide any specific details about the hardware used for the experiments, such as CPU/GPU models or memory specifications.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x).
Experiment Setup Yes Latent factor dimension K = 64, regularization factor λ = 0.00005 and learning rate is 0.05. In TAPITF, d = 0.5, time unit is day. Latent factor dimension and regularization factor is the same as PITF. The iteration number of PITF and TAPITF are both 100.