Model generation system
The model generation system addresses the challenge of obtaining learning data by generating a recommendation condition derivation model to estimate user movement and extract optimal conditions, facilitating accurate simulation and determination of recommendation conditions that achieve target values.
JP7868251B2Active Publication Date: 2026-06-01NTT DOCOMO INC
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- NTT DOCOMO INC
- Filing Date
- 2023-12-06
- Publication Date
- 2026-06-01
Smart Images
Figure 0007868251000001 
Figure 0007868251000002 
Figure 0007868251000003
Abstract
To generate a recommendation condition derivation model, this model generation system is provided with: a recommendation condition candidate generation unit that acquires a plurality of recommendation condition candidates; a user movement estimation unit that acquires user movement estimation information on the basis of each of the recommendation condition candidates and transition of positions of users indicated by movement demand information; a recommendation condition extraction unit that extracts, as an optimum recommendation condition, a recommendation condition candidate with the largest degree of achievement for a given target value on the basis of the user movement estimation information for each recommendation condition candidate; and a recommendation condition derivation model training unit that performs machine learning of a recommendation condition derivation model using training data with target information, attribute information relating to each user, and user position information as explanatory variables and the optimum recommendation condition as an objective variable.
Need to check novelty before this filing date? Find Prior Art