: Research has explored how organizing results into dynamic categories or clusters can be more helpful for general questions compared to standard relevance rankings. Studies also indicate that users often trust and select top-ranked results even if they are less relevant or credible.
Quantifies how past user data improves results for rare, complex queries.
“I recommend looking for the most recent article on your topic, since the literature cited... will also be the most recent.” UC Santa Barbara · 3 weeks ago
Compares relevance patterns between major search engines like Google and Bing.
: Significant research identifies a positive relationship between past user data and search engine quality. For example, studies show that while new entrants can match established engines for popular queries, having a history of user-generated data is critical for providing high-quality results for "rare queries".
Finds that explaining why a result was ranked increases user trust and efficiency. Academic Perspectives on Search Quality
Investigates methods to reduce topical bias and improve result diversity.
Examines how grouping results helps users explore general topics.