When specifying metadata for a collection, especially the name or description of a program, study, or asset, consider the perspective of users who will be searching for that collection. If the name of your collection contains the same words as many other collections in MoDaC, it will be more difficult for the users to find your collection. In each metadata field, specify what makes your collection unique. Spell out any acronyms and avoid jargon.
| Metadata | Guidelines |
|---|---|
| Collection Names | For each study, capture the broad effort. For each asset, indicate the sub-task that addresses a portion of the study. If the asset title is the same or very similar to the study title, then it means that the asset encompasses the whole of the study (in which case, it may be better to have the asset listed under a broader existing study than creating one for it). |
| Collection Descriptions | Provide a detailed description of the purpose and function of your resource so that an outside researcher will understand the significance of your resource, how they might use it, and what the resource can help them do. |
For more details, refer to the following pages:
The following article clarifies what is important in a model:
Harish KB, Price WN, Aphinyanaphongs Y. Open-Source Clinical Machine Learning Models: Critical Appraisal of Feasibility, Advantages, and Challenges. JMIR Form Res. 2022 Apr 11;6(4):e33970. doi: 10.2196/33970. PMID: 35404258; PMCID: PMC9039816 
.
Add Comment