Nonparametric-Dirichlet-process-based-Latent-Factor-Model

citation

If you use the code or datasets in this repository, please acknowledge the paper "Fine-Grained Job Salary Benchmarking with a Nonparametric Dirichlet-process-based Latent Factor Model"

data

You can download the raw datasets used in this paper via the link (cleaned, in csv format, with data descriptions):

https://drive.google.com/drive/folders/1eW_30c-ZHYfBip4GAed-DsG9OLqyJZXT?usp=sharing

You can download the preprocessed data in this paper via the link (in pickle format, for code running purpose):

https://drive.google.com/drive/folders/1RsTLFg5tSmESQAkypr5G8-tW48nGMO66?usp=sharing

Please be aware that if you want to run the model using our datasets directly, you need to download the preprocessed data and put them under the directory: "src/source_file" .

prerequisites (please install the following packages before you run this model NPD-JSB)

python 3.6
numpy 1.19.2
pandas 1.1.3
scipy 1.5.2

instructions to run the model NPD-JSB

The src folder includes the source code of the model NPD-JSB.

To run the model, you need to download the whole src folder. Then open a terminal to run the commands as below. According to the different data splitting settings and datasets, there are four different options. I list the four options and the corresponding commands as below.

option 1 (running the NPD-JSB model on the IT dataset with 5 cross-validations):

python main.py --data_split_type cross_validate --data_source it

option 2 (running the NPD-JSB model on the IT dataset with varying proportional data splittings):

python main.py --data_split_type proportional --data_source it

option 3 (runing the NPD-JSB model on the Finance dataset with 5 cross-validations):

python main.py --data_split_type cross_validate --data_source finance

option 4 ((runing the NPD-JSB model on the Finance dataset with varying proportional data splittings):

python main.py --data_split_type proportional --data_source finance

appendix

The appendix.pdf is the online supplement file for the paper "Fine-Grained Job Salary Benchmarking with a Nonparametric Dirichlet-process-based Latent Factor Model". It includes:

Proof of the Variational Inference Process,
Proof of Updating Formulas,
The Algorithms of Generative and Optimization Process of NDP-JSB,
Baselines Introduction,
Statistics and Descriptions of Datasets,
Additional Experimental Results.

experimental results

You can access more results from the oringinal paper.