You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The node normalizer now returns an information content for some nodes. One way this might be used is by ARAs that decide to prioritize paths going through high IC nodes or dropping paths going though low-IC nodes. It could be a build time, or used as a component in ranking.
KPs might also use this information to create or modify edge scores. So a KP might calculate an edge evidence score, and it might decide to use IC as a component int hat calculation.
Do we need to be concerned about multiple, possibly conflicting uses? If a KP uses IC in a score, are we at risk of double counting or over-emphasizing?
What about other similar properties? Do we need rules about when they are used, or provenance saying what was used?
The text was updated successfully, but these errors were encountered:
The node normalizer now returns an information content for some nodes. One way this might be used is by ARAs that decide to prioritize paths going through high IC nodes or dropping paths going though low-IC nodes. It could be a build time, or used as a component in ranking.
KPs might also use this information to create or modify edge scores. So a KP might calculate an edge evidence score, and it might decide to use IC as a component int hat calculation.
Do we need to be concerned about multiple, possibly conflicting uses? If a KP uses IC in a score, are we at risk of double counting or over-emphasizing?
What about other similar properties? Do we need rules about when they are used, or provenance saying what was used?
The text was updated successfully, but these errors were encountered: