Open uschindler opened 6 years ago
The current PANGAEA recommender offers recommendation (siliarity based one) using title of dataset, related paper title, (P)ID of platform/basis/ship, (P)ID of authors, (P)ID of campaign, scientific topic (from ontology), geographical location, and datetime. To improve this, the PID graph can be used to also look into information more for away: e.g., author of dataset is scientist in a working group / institution (retrieved from ORCID) and you want to present datasets by other people from same working group as "related". The recommender is online since August, 6, 2018. It already uses the "PID graph" as it uses identifiers and builds relationships using them, but it does not dive deeper (as described above). So we can already use it as an example as-is.
This issue was moved to https://www.pidforum.org/t/data-recommender/104
As a user of PANGAEA's data portal, I'd like to get dataset recommendations, so i can find related datasets, covering the the same area of interest. The current recommender on PANGAEA cannot present recommendations about datasets measured with similar instuments, but that's something I'd like to look into.