In our study, we developed a decision support system for document retrieval and study selection.
In order to run this code, clone our repository, and add an 'embeddings' and 'weights' folder to the root. Furthermore, place the 100-dimensional GloVe embeddings from GloVe6B.zip in the embeddings folder (https://nlp.stanford.edu/projects/glove/).
Upload the code to Google Drive, and open the iPython Notebook in Google Colab. Then, you can perform your SLR study. Please email to raymon.vandinter@wur.nl if you have any questions. Do not hesitate to file an issue or perform a pull request.
The tutorial video shows two examples. the second example is at around 4 minutes.
When citing this code, please use van Dinter, R., Catal, C., & Tekinerdogan, B. (2021). A Decision Support System for Automating Document Retrieval and Citation Screening. Expert Systems with Applications, 115261.