Abstract
Some search engines provide query recommendations that suggest additional queries related to a user’s original query. Low-quality query recommendations can be irrelevant to the user’s information need and degrade the user experience. This disclosure describes techniques to filter low-quality query recommendations such that they are not displayed to users. The filtering is performed by evaluating a set of candidate query recommendations based on query recommendability scores along three quality-related dimensions - content, formulation, and flavor. Additionally, updateworthiness for a query recommendation can also be used. Implementation of the techniques can raise the quality, utility, and trustworthiness of query recommendations and make it more efficient and effective for users to obtain relevant information, thus enhancing the search user experience.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Jolly, Jap Leen Kaur; Rai, Karan; Shroff, Paritosh; and Saade, Stephanie, "Filtering Low Quality Query Recommendations", Technical Disclosure Commons, (March 30, 2025)
https://www.tdcommons.org/dpubs_series/7948