Saifuddin Kharawala, Anadi Mahajan and Paul Gandhi
In line with recent publications, our in-house research has identified some potential for the application of NLP technology for title and/or abstract screening for SLRs, although limitations remain.
We found that NLP technology was inadequate for full text screening and extractions from manuscripts.
If journals were to mandate key data from published studies be made available in structured formats, alongside manuscripts, this might open up the scope for MLand/or NLP techniques for automated screening and extraction
