Predicting a consumer selection preference based on estimated preference and environmental dependence

• US 10,255,560 B2
• Filed: 09/07/2017
• Issued: 04/09/2019
• Est. Priority Date: 08/21/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Katsuki, Takayuki, Osogami, Takayuki
• Primary Examiner(s): Vincent, David R
• Application Number:
  US15/698,153
• Publication Number:
  US 20170372234A1
• Time in Patent Office: 579 Days
• Field of Search: 706 15, 706 45, 706 12
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

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.

Claims

1. An information processing method executed by a computer, the method comprising:

acquiring, by a processor, a feature value for a feature possessed by each of a plurality of selection objects;

acquiring history data including a history indicating that a plurality of selection subjects have selected selection objects;

learning processing by allowing 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 for 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,

wherein the choice model calculates the selectability, with which each selection subject selects each selection object in each selection environment, based on a product of the feature vector of the selection object and the preference vector of the selection subject, and an element corresponding to the selection object in the environment-dependent vector corresponding to the selection environment,

wherein each of elements of the preference vector of each selection subject and the environment-dependent vector in each selection environment is represented by a prior distribution;

calculating distribution parameters of the prior distribution on each of the elements of the preference vector of each selection subject and the environment-dependent vector in each selection environment by learning;

generating distribution parameters of prior distributions of the preference vector of each selection subject and the environment-dependent vector in each selection environment;

generating a next sample of the environment-dependent vector in each selection environment based on the prior distribution of the environment-dependent vector in each selection environment;

generating a next sample of the preference vector of each selection subject based on the prior distribution of the preference vector of each selection subject;

calculating distributions of the environment-dependent vector in each selection environment and the preference vector of each selection subject by a Bayesian estimation based on a Markov chain Monte Carlo method (MCMC) of the samples of the environment-dependent vector in each selection environment and the preference vector of each selection subject that occur multiple times;

simulating learning results obtained in the learning processing; and

outputting the simulation of the learning results on a display.