ISTEX_MentalRotation use-case of SSbE

Repeatability code for SSbE paper of SERecSys Workshop of ICDM:

To start, you would need to generate the dataset using the script inside IstexDataDownload_Treatment folder (you may also dowload it from: https://drive.google.com/drive/folders/1i1I3fi6Qgdz-A4hI8_MkjZCg-EPieTi_?usp=sharing)
The main file to start with is bow_svd.py. It will transform the whole corpus into its semantic features representation.
Then, run classifier.py to train and to generate ranked results.
The baseline is using More_Like_This Query of ElasticSearch.
You can find the users annotations in the annotations folder. The notebook comparatrive_evaluation.ipynb provide the initial evaluation
For active learning process, you should open and run the cells of build_dataset_active_learning.ipynb
You may also check other available notebooks for further analysis. Other .py files like LDA for topic analysis are also available

Name		Name	Last commit message	Last commit date
Latest commit History 55 Commits
IstexDataDownload_Treatment		IstexDataDownload_Treatment
annotations		annotations
elasticsearch		elasticsearch
results		results
sample_data		sample_data
utils		utils
CompareSVDVectors.py		CompareSVDVectors.py
IstexDoc2Vec.py		IstexDoc2Vec.py
LDA.py		LDA.py
License		License
README.md		README.md
SSbE_Presentation_SERecSys_ICDM_18Nov2017.pdf		SSbE_Presentation_SERecSys_ICDM_18Nov2017.pdf
Sample4LDA.py		Sample4LDA.py
TrendingDataForChart.py		TrendingDataForChart.py
bow_svd.py		bow_svd.py
classifier.py		classifier.py
comparative_evaluation.ipynb		comparative_evaluation.ipynb
ids2docs.py		ids2docs.py
ids2docs_years.py		ids2docs_years.py