Default Rating Estimation Using Poisson Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing default rating methods in collaborative filtering systems fail to accurately maintain a user's rating profile when default ratings are added to sparse datasets, as they do not consider the distribution of user and item ratings effectively.
Innovation Solution
A method using a Poisson distribution to estimate default ratings, which involves calculating a λ value based on the user's average rating and weighting it with item and dataset averages to ensure the default rating reflects the rating distribution, thereby maintaining the user's rating profile and preventing dataset flattening.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional default rating methods are used in collaborative filtering systems, then the system can handle sparse datasets, but the user's rating profile is not accurately maintained and dataset flattening occurs
Solution Approach 1:
The patent transforms the rating estimation problem by changing the parameter representation from direct rating values to Poisson distribution parameters (λ). Instead of estimating a single default rating value, the system models rating frequencies using Poisson distribution, where λ is derived from user average ratings. This parameter transformation enables probabilistic modeling that preserves user rating profiles while handling sparsity, resolving the contradiction between reliability and complexity.
Solution Approach 2:
The patent introduces Poisson distribution as an intermediary statistical model between the sparse rating data and the default rating estimation. The distribution acts as a mediator that captures the probabilistic nature of user ratings, allowing the system to infer default ratings while maintaining the underlying rating profile characteristics. This intermediary model prevents direct copying of average ratings that would cause dataset flattening.
2Quantity of substance
If default ratings are added to sparse datasets using conventional methods, then more complete rating data is obtained, but the rating distribution becomes flattened and loses user-specific characteristics
Solution Approach 1:
The patent adds a probabilistic dimension to the rating estimation by introducing Poisson distribution parameters. Instead of filling missing ratings with scalar values from conventional methods, the system uses λ parameters that encode both the magnitude and distribution characteristics of user ratings. This dimensional enrichment allows complete rating matrices to be generated while preserving user-specific rating profiles through the probabilistic structure.
Solution Approach 2:
The system changes the parameter from direct rating values to Poisson distribution parameters (λ), where λ is calculated based on user average ratings. This parameter transformation ensures that when default ratings are generated, they reflect the user's rating behavior pattern rather than simply copying global averages, thus maintaining precision in rating profile representation while increasing data completeness.
3Ease of operation
If user average ratings are directly used as default ratings, then the estimation process is simple, but it does not reflect the rating distribution and biases results
Solution Approach 1:
The patent replaces the mechanical arithmetic operation of directly using average ratings with a probabilistic substitution approach. Instead of simply copying the user average rating value, the system uses the average to derive the Poisson parameter λ, which then generates ratings according to a probability distribution. This substitution maintains computational simplicity while dramatically improving the accuracy of rating distribution reflection by accounting for variance and user-specific patterns.
Data Source
AI summary
A method of estimating a default rating of a rated dataset is provided, where the dataset comprises at least one series of ratings associated with at least one user and each series comprise ratings associated with at least two items. For a reference user and an item for which a rated value is missing the item's average rating, ir, the reference users average rating, Ru, and the datasets average rating, dr, is collected. A Poisson distribution of the reference users rating is then generated on the basis of the reference users average rating. A random Poisson rating, ur, is calculated on the basis of the Poisson distribution, and a default rating, r, is estimated by weighting the random Poisson rating on the basis of the items average rating, the users average rating and the datasets average rating.