conceptinterest-graph-recs
Explicit vs implicit recommendation
Notes
Explicit vs implicit recommendation
One-line summary: A literature claim in Roth et al. (2020): filter bubbles are generally observed more with user-declared (explicit) recommenders than with activity-based (implicit) ones. Tension for a user-owned interest graph, which is an explicit object.
The insight
A user-drawn interest graph is an explicit preference store. Roth, Mazières, and Menezes’s PLoS ONE abstract says recent empirical studies generally locate filter bubbles more in explicit recommendation than in implicit recommendation. That sentence is a pointer to other work, not a test of a hand-built YouTube graph. It cuts against the hope that drawing the graph automatically centers the feed. From 2026-08-21-academic-research-user-owned-interest-graphs-vs-platform-recs.
Evidence
- From 2026-08-21-academic-research-user-owned-interest-graphs-vs-platform-recs: “filter bubbles may principally be observed in the case of explicit recommendation (based on user-declared preferences) rather than implicit recommendation (based on user activity).”
- From 2026-08-21-academic-research-user-owned-interest-graphs-vs-platform-recs: “That sentence is a literature claim in their abstract, not a test of a user-drawn interest graph.”
- From 2026-08-21-academic-research-user-owned-interest-graphs-vs-platform-recs: UCRS / LACE / learner-PKG abstracts treat explicit control as a fix for over-personalization — “Both can be true if ‘explicit’ means a static declared profile versus an editable control surface.”
Contradictions / tensions
- Roth’s literature pointer vs user-control papers (user-controllable-recommender). This pass did not retrieve the studies Roth alludes to. Tension stays open.
Open questions
Related
Referenced by