A GenAI Approach to Suggesting Broad and Narrow Queries for Exploratory Search
Hanieh Khakshoor, Orland Hoeber · Proceedings of the Association for Information Science and Technology · 2025
ABSTRACT Query reformulation is a critical aspect of the exploratory search model. During exploratory browsing, searchers issue broad queries to learn about the breadth of the topic; during focused searching, they issue narrow queries to learn about a specific aspect of the topic. Typically, little support is provided by the search interface to help searchers refine and reformulate their queries. This research addresses this gap, demonstrating a generative AI approach to suggesting both broad and narrow queries. Using a timeline of past search activities (queries issued and search results saved), the approach allows searchers to reformulate their queries based on specific saved search results, previous queries issued, or the current query. This gives the searcher interactive control not only over whether to request broad or narrow query suggestions, but also upon what information to base those suggestions.