US 12,229,202 B2
Inspiring divergent thinking and creativity via curiosity halo
Joseph Anthony Russo, Westford, MA (US); Jennifer Anne Janechek, North Liberty, IA (US); Jill Sherman, Northfield, IL (US); Jonathan Ralph, Rosemount, MN (US); and Christopher P Wilczak, Chicago, IL (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jun. 26, 2023, as Appl. No. 18/214,114.
Prior Publication US 2024/0427825 A1, Dec. 26, 2024
Int. Cl. G06F 16/9532 (2019.01); G06F 16/9535 (2019.01); G06F 16/954 (2019.01)
CPC G06F 16/9532 (2019.01) [G06F 16/9535 (2019.01); G06F 16/954 (2019.01)] 16 Claims
OG exemplary drawing
 
1. A computer implemented recommender method comprising:
accessing, by a user, via a remote computer system, an on-line recommender application used to generate recommendations on a given search topic or concept;
selecting, by a hardware processor, based on a user profile information, a machine-learned recommender model for use by the recommender application to search web-based platforms and generate search results based on the given topic or concept, the selected recommender model responsive to an initial recommendation threshold parameter defining an initial search results boundary of content type generated when conducting an initial search query of the given topic or concept, said initial search results boundary based upon said user profile information, wherein content provided in response to search queries based on the initial recommendation threshold parameter comprise initial search results aligned with the user profile information;
receiving, via an interface at the remote computer system, a spectrum range model parameter defining an expanded search results boundary of content type that includes diversified content beyond the types of content generated within the initial search results content boundary;
running, at the hardware processor, the recommender model responsive to the spectrum range model parameter for conducting a search query and generating a halo of recommendation results, the generated halo of recommendation results comprising a search results dataset having an expanded and diversified content range for the given topic or concept to a degree based upon an application data set and the expanded search results boundary, said hardware processor subtracting, from the initial search results, generated results aligned with the user profile information that the user has already been exposed to, wherein said halo of recommendation results displays expanded and diversified content the user has not yet seen; and
displaying, via the user interface of the remote computer system, the generated halo of recommendation results having the expanded and diversified content range generated by said recommender model that the user has not yet seen.