17535558. CREATIVITY METRICS AND GENERATIVE MODELS SAMPLING simplified abstract (International Business Machines Corporation)
Contents
CREATIVITY METRICS AND GENERATIVE MODELS SAMPLING
Organization Name
International Business Machines Corporation
Inventor(s)
Payel Das of Yorktown Heights NY (US)
Brian Leo Quanz of Yorktown Heights NY (US)
Skyler Speakman of NAIROBI (KE)
Victor Abayomi Akinwande of Karen (KE)
CREATIVITY METRICS AND GENERATIVE MODELS SAMPLING - A simplified explanation of the abstract
This abstract first appeared for US patent application 17535558 titled 'CREATIVITY METRICS AND GENERATIVE MODELS SAMPLING
Simplified Explanation
Abstract: The patent application describes a method for estimating creativity scores for given samples and filtering them based on these scores. The method involves initializing gaussian distributions with random parameters for creative scores, non-creative scores, and normal scores. These parameters are then updated using gaussian mixture models with expectation maximization to cluster the different types of scores. The creativity score for each sample is estimated based on the probability of it belonging to a creative cluster. The method filters the samples based on their creativity scores to generate a set of optimal samples.
Patent/Innovation:
- Gaussian distributions are initialized with random parameters for creative scores, non-creative scores, and normal scores.
- Random parameters are updated using gaussian mixture models with expectation maximization to cluster the different types of scores.
- Creativity scores are estimated for given samples based on the probability of them belonging to a creative cluster.
- Samples are filtered based on their creativity scores to generate a set of optimal samples.
Potential Applications:
- Creative content generation in fields such as art, music, writing, etc.
- Identifying and selecting innovative ideas or solutions in various industries.
- Personalized recommendation systems for creative products or experiences.
- Assessing and comparing creativity levels in individuals or groups.
Problems Solved:
- Difficulty in quantifying and measuring creativity objectively.
- Lack of efficient methods to filter and select creative samples from a large dataset.
- Subjectivity in evaluating and identifying creative content or ideas.
- Inability to estimate creativity scores for given samples accurately.
Benefits:
- Provides a systematic and objective approach to estimate creativity scores.
- Enables efficient filtering and selection of creative samples from a dataset.
- Reduces subjectivity in evaluating and identifying creative content or ideas.
- Facilitates personalized recommendations and assessments based on creativity levels.
Original Abstract Submitted
A gaussian distribution is initialized with random parameters for each of creative scores, non-creative scores, and normal scores. The random parameters are updated to maximize a likelihood of each sample score from a training set being from its corresponding gaussian distribution by using one or more gaussian mixture models with expectation maximization to cluster the creative scores, the non-creative scores, and the normal scores. A creativity score is estimated for each of a plurality of given samples as a probability of the corresponding given sample being in a creative cluster given one or more of the random parameters of a corresponding gaussian distribution. One or more of the plurality of samples are filtered based on the creativity scores to generate a set of optimal samples.
