If a list is generated manually, it's easy to track changes of the list by "view history" function. But if the list is generated automatically, it's not easy to do so.

If there were only triple additions (not removals), it's easy to track changes in this way:

1. Create a new table listing all RDF triples and the revision which some triple is added.
2. When querying edits affecting a specific query, we may found all related triples and related revisions. Then we remove any duplicate revisions and sort the list of revisions by date. We got the result we needed.

Open question:

1. How to deal with triple removals? Should we create a new table for all former triples? and how to query them?
2. How to integrate it and traditional article history?