Choice Model for Consumer Preference Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional conjoint analysis using logit models fails to accurately predict consumer preferences as it does not consider environmental influences such as cognitive biases and point-of-sale situations, leading to imprecise preference estimation.
Innovation Solution
An information processing apparatus and method that learns a choice model using history data to incorporate both feature preferences and environmental dependencies, allowing for the calculation of selectability of selection objects across various environments, thereby considering environmental factors like cognitive biases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional logit model is used for conjoint analysis, then the model structure is simple and easy to implement, but the prediction accuracy of consumer preferences deteriorates due to ignoring environmental influences
Solution Approach 1:
The patent segments the selection process into two distinct components: (1) preference for selection objects based on their features, and (2) environmental dependence affecting ease of selection. This segmentation allows the model to separately learn and combine these two aspects, improving prediction accuracy while maintaining a structured approach that balances complexity.
Solution Approach 2:
The patent introduces a new dimension - environmental dependence - to the traditional logit model. By adding this dimension that captures contextual factors (cognitive biases, point-of-sale situations), the model transitions from considering only feature-based preferences to also incorporating environmental influences, thereby improving preference estimation accuracy.
2Adaptability or versatility
If conventional logit model is used, then the model only considers feature preferences, but it fails to account for environmental factors such as cognitive biases and point-of-sale situations
Solution Approach 1:
The patent makes the model dynamic by introducing environmental dependence that varies across different selection environments. Instead of a static preference model, the system now adapts to different contexts (e.g., different stores, different display conditions) by learning environment-specific selection patterns, thereby improving both adaptability and estimation accuracy.
Solution Approach 2:
The patent introduces environmental dependence as an intermediary factor that mediates between the selection subject and the selection objects. This intermediary captures the influence of environmental factors (cognitive biases, point-of-sale situations) on the selection process, allowing the model to account for these factors without directly modeling each environmental variable.
3Measurement precision
If conventional logit model is used, then the learning process is simple, but it cannot distinguish between true preference and environmental influence on selection
Solution Approach 1:
The learning process is segmented into two parallel learning tasks: learning preference for selection objects and learning environmental dependence. This segmentation allows the system to efficiently learn both components separately using historical data, then combine them for accurate preference measurement without requiring excessively long learning periods.
Solution Approach 2:
The patent merges the learning of preference and environmental dependence into a unified framework that leverages historical selection data. By combining these two learning processes and integrating their outputs through the selectability calculation, the system achieves accurate preference measurement while utilizing existing data efficiently.
Data Source
AI summary
An information processing apparatus includes a history acquisition section configured to acquire history data including a history indicating that a plurality of selection subjects have selected selection objects; a learning processing section configured to allow a choice model to learn a preference of each selection subject for a feature and an environmental dependence of selection of each selection object in each selection environment using the history data, where the choice model uses a feature value possessed by each selection object, the preference of each selection subject for the feature, and the environmental dependence indicative of ease of selection of each selection object in each of a plurality of selection environments to calculate a selectability with which each of the plurality of selection subjects selects each selection object; and an output section configured to output results of learning by the learning processing section.