When a user is starting to annotate a new metadata record via https://discovery.biothings.io/guide/outbreak/dataset, it'd be cool to suggest that instead a user edit the previously entered metadata, if that dataset has already been added.
The problem: if we open up /guide to the public, I imagine there will be a lot of duplication of key datasets, such as the JHU epidemiology dataset, or the WHO Situation Reports. We don't want to waste the curator's time, nor do we want to add redundant data.
Suggested solution: as the user is entering the required fields, the front-end checks the Resources API for any similar dataset and pops up a suggestion that instead of creating a new metadata record, they can edit the existing one. Similar to how Stack Overflow suggests that a question has already been answered, or any number of online chat customer service forums.
Will require some sort of ML to predict the dataset a user is intending to add; would prioritize authors, name, description, abstract.
When a user is starting to annotate a new metadata record via https://discovery.biothings.io/guide/outbreak/dataset, it'd be cool to suggest that instead a user edit the previously entered metadata, if that dataset has already been added.
The problem: if we open up /guide to the public, I imagine there will be a lot of duplication of key datasets, such as the JHU epidemiology dataset, or the WHO Situation Reports. We don't want to waste the curator's time, nor do we want to add redundant data.
Suggested solution: as the user is entering the required fields, the front-end checks the Resources API for any similar dataset and pops up a suggestion that instead of creating a new metadata record, they can edit the existing one. Similar to how Stack Overflow suggests that a question has already been answered, or any number of online chat customer service forums.
Will require some sort of ML to predict the dataset a user is intending to add; would prioritize authors, name, description, abstract.