reach
reach copied to clipboard
wellcometrust
Wellcome tool to parse references scraped from policy documents using machine learning
Instead of separating the two search functions, we should explore the option of displaying only a single search to users with advanced search and filtering functionality added before the run...
https://github.com/wellcometrust/datalabs/pull/523 At the moment we achieve a F1 of 0.55 in finding people entities.
@aoifespenge commented on [Thu Jul 18 2019](https://github.com/wellcometrust/datalabs/issues/362) --- @nsorros commented on [Fri Jul 19 2019](https://github.com/wellcometrust/datalabs/issues/362#issuecomment-513283213) To a certain extent that is a feature, as we have decided to filter out...
Written by Alex Mankoo last year. I think we are address this in the product but we should consider this anyway. Potential actions to deal with this: 1) Identify how...
@aoifespenge commented on [Wed Feb 26 2020](https://github.com/wellcometrust/datalabs/issues/607)
@aoifespenge commented on [Wed Feb 26 2020](https://github.com/wellcometrust/datalabs/issues/608)
This is a fairly sizeable refactor target which covers a few different issues. **Warning** Before anyone starts on this, this issue might be negated by architectural changes proposed in #419...
This is nitpicky, but it cost me some time today. file_hash/document_id is stored in two places in the dict output by `ExtractRefs`: `document_id` and `metadata['file_hash']`. The information is the same,...
Currently in the 'Discover Citations' results page, users can see part or all of the research publication title. We believe that users (desk researchers) will want a link to the...
Spotted by Alex: After a couple of different search terms it seems like the Publication Year and Document year are same, when this is unlikely the case, as usually it...
Metadata
Owner
Metadata
Wellcome tool to parse references scraped from policy documents using machine learning