Fusing Social Networks with Deep Learning for Volunteerism Tendency Prediction
Authors: Yongpo Jia, Xuemeng Song, Jingbo Zhou, Li Liu, Liqiang Nie, David Rosenblum
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | With extensive experimental evaluations, we demonstrate the effectiveness of our model, which outperforms several state-of-the-art approaches in terms of precision, recall and F1-score. |
| Researcher Affiliation | Collaboration | 1NUS Graduate School for Integrative Sciences and Engineering, National University of Singapore, Singapore 2School of Computing, National University of Singapore, Singapore 3Big Data Lab, Baidu Research, China {jiayongpo, dcsliuli}@nus.edu.sg, {sxmustc, nieliqiang}@gmail.com, jzhou@baidu.com, david@comp.nus.edu.sg |
| Pseudocode | No | The paper describes the optimization process using mathematical equations but does not include any pseudocode or algorithm blocks. |
| Open Source Code | No | The paper only provides a link to the compiled dataset, not to the source code for the methodology or experiments. 'The compiled dataset is currently publicly available via: http://multiplesocialnetworklearning.azurewebsites.net/' |
| Open Datasets | Yes | The compiled dataset is currently publicly available via: http://multiplesocialnetworklearning.azurewebsites.net/ |
| Dataset Splits | Yes | To avoid overfitting and achieve the best performance, we selected the optimal parameters for each model based on 10-fold cross validation, and we performed another 9-fold cross validation on the training data with grid search in each round (i.e. nested cross-validation). Hence, in each experiment, for each round of the 10-fold cross validation, 90% of the samples were used for training the model with 9-fold cross validation, and the remaining 10% were reserved for testing. |
| Hardware Specification | No | The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU/CPU models, memory). |
| Software Dependencies | No | The paper mentions 'Python' and 'Theano' but does not specify version numbers for these or any other software dependencies needed for reproducibility. |
| Experiment Setup | Yes | To save the cost of memory and computation, we use three hidden layers to construct our FARSEEING model and its variants as well as the DBN and M-DBM, which is sufficient to achieve good performance. ... For the grid search, it was conducted between 10 2 and 102 with small but adaptive step sizes. The step sizes are 0.01, 0.05, 0.5 and 5 for the range of [0.01, 0.1], [0.1, 1], [1, 10] and [10, 100], respectively (Nie et al. 2015). |