Extrapolating Interaction Data Using Mixed Granularity Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating the performance of interactive computing environments are incomplete due to the lack of individual interaction data, leading to an aggregated impact evaluation that does not accurately reflect user experiences.
Innovation Solution
The application of machine learning models, specifically a mixed granularity model and attribution models, to extrapolate aggregated interaction data and identify the impact of individual actions leading to a target action, enabling modifications to user interfaces for improved user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If aggregated interaction data is used for evaluation, then data privacy is protected and data aggregation is simplified, but evaluation precision is reduced due to loss of individual user behavior details
Solution Approach 1:
The patent creates synthetic copies of individual user interaction sequences by extrapolating from aggregated data using machine learning models. These synthetic sequences replicate the behavior patterns of individual users without requiring actual individual data, thus maintaining privacy while enabling precise evaluation of individual user journeys and their impact on conversion goals.
2Measurement precision
If individual interaction data is collected for precise evaluation, then evaluation precision is improved, but data privacy risks increase and data collection complexity increases
Solution Approach 1:
The patent introduces an intermediary layer between data collection and evaluation. Aggregated data serves as the input, and machine learning extrapolation models act as the intermediary that transforms this aggregated data into synthetic individual-level sequences. This intermediary enables precise individual user evaluation without requiring direct collection of sensitive individual data, thus reducing privacy risks and collection complexity.
3Productivity
If aggregated impact evaluation is performed, then evaluation process is simplified, but accuracy of user experience assessment is reduced
Solution Approach 1:
The patent performs preliminary action by pre-processing aggregated data through machine learning models to generate synthetic individual user sequences before the actual evaluation takes place. This preliminary transformation enables the evaluation system to treat aggregated data as if it were individual data, maintaining both the simplicity of working with aggregated data and the accuracy of individual user experience assessment.
Data Source
AI summary
In some embodiments, a computing system extrapolates aggregated interaction data associated with users of an online platform by applying a mixed granularity model to generate extrapolated interaction data for each user in the users. The aggregated interaction data includes a total number of occurrences of a target action performed by the users with respect to the online platform. The extrapolated data includes a series of actions leading to the target action for each user. The computing system identifies an impact of each action in the series of actions for each user on leading to the target action based, at least in part, upon the extrapolating a series of actions associated with the user. User interfaces presented on the online platform can be modified based on at least the identified impacts to improve customization of the user interfaces to the users or enhance an experience of the users.