New Cross Dataset Query Interface!

2 posts
Was this reply useful? Learn more...
   
[-] licongcui +0 points · almost 3 years ago

The new Physio-MIMI cross dataset query interface for NSRR is available on https://informatics.case.edu. The goal of the interface is to be directly used by sleep researchers, for activities across multiple datasets such as data exploration for deriving new hypotheses as well as for cohort identification. There are two main features:

  1. Query Builder, which provides dynamic construction of advanced cross-cohort queries, with terminology support and visual controls such as slider bar, radio buttons, and checkboxes;

  2. Query Manager, which allows for storing and labeling queries for reuse.

If you have any questions or suggestions, please contact us at nsrrphysiomimi@gmail.com. We look forward to hearing from you so that we can provide more interface features and improve the interface usability. Thanks!

109 posts
Was this reply useful? Learn more...
   
[-] mrueschman +0 points · almost 3 years ago

Licong,

Nice work!

I had a couple questions:

  1. Is there any way to construct a query with OR statements? (It seems like I can only build with AND statements right now.)
  2. When running a query within a single dataset (e.g. SHHS), would it be possible to get the "combined" count as well? For instance, I searched for RDI3P between 15-30 and and AGE between 60-65 in the SHHS dataset, yet I did not receive the count for where these intersect. That is, how many people between the age of 60-65 had an RDI between 15-30?

Thanks!

2 posts
Was this reply useful? Learn more...
   
[-] licongcui +0 points · almost 3 years ago

Hi Mike,

Thanks for the questions.

  1. Right now the cross dataset query interface only supports AND statements among core terms. OR statements are available for the options provided for a core term or the variables mapped to a core term. Including OR statements among core terms is not a technical issue, but might cause confusion or difficulty to novice users. Indeed, this is a balance between simplicity and complexity of the interface.

  2. I am able to get the "combined" count within a single dataset using the example you provided, and cannot reproduce the issue. Please try again and let me know if you still experience such issues.

Thanks, Licong

Write a Reply