Enriching Activity Data Records with Entity-Identifying Characteristics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
User activity data records often lack clarity, leading to incorrect identification of activities as fraudulent, due to incomplete or unclear data, resulting in false fraud claims.
Innovation Solution
A computer-based system that utilizes machine-learning modeling to enrich activity data records by incorporating entity-identifying characteristics from other users' data, using a mobile device to gather additional user-provided data such as text descriptions, images, or categorizations, and updates the records to prevent false fraud claims.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If activity data records are stored with limited electronic sources, then storage efficiency is improved, but data clarity and completeness deteriorate
Solution Approach 1:
The system merges activity data from multiple electronic sources including mobile device data, social media platforms, and third-party applications to create enriched activity records. This combination of diverse data sources resolves the contradiction by maintaining storage efficiency while significantly improving data clarity and completeness through integrated information from multiple origins.
Solution Approach 2:
The system introduces an intermediary enrichment process that acts as a mediator between raw activity data and final activity records. This intermediary layer adds contextual information, entity identifications, and descriptive details without requiring permanent storage of all raw data, thus maintaining storage efficiency while improving data clarity.
2Loss of information
If activity data records are enriched with additional data from multiple sources, then data clarity is improved, but device complexity increases
Solution Approach 1:
The system segments the data enrichment process into distinct modular components: data collection module, data processing module, entity identification module, and record updating module. Each component handles specific tasks independently, which reduces overall system complexity by making each segment manageable and independently optimizable while achieving comprehensive data enrichment.
Solution Approach 2:
The system implements self-service mechanisms where the enrichment process automatically retrieves and processes data from multiple sources without requiring manual intervention. The system autonomously identifies entities, enriches records, and updates databases, reducing operational complexity while maintaining high data clarity.
3Measurement precision
If machine learning models are trained on user data to identify entities, then entity identification accuracy is improved, but loss of time for data processing increases
Solution Approach 1:
The system performs preliminary actions by pre-training machine learning models on historical user data and pre-processing activity data to extract relevant features before entity identification is needed. This advance preparation significantly reduces real-time processing time while maintaining high entity identification accuracy, as the heavy computational work is completed beforehand.
Solution Approach 2:
The system applies partial action by using machine learning models only for specific entity identification tasks rather than processing all data uniformly. The model focuses on identifying key entities and characteristics that are most important for activity record enrichment, reducing overall processing time while maintaining high accuracy for critical identification tasks.
Data Source
AI summary
Systems and methods are disclosed including determining that a first activity data of a completed activity record of a user comprises a predictive characteristic indicative of a potential rejection claim. The computing device produces a request for a second activity data of the completed activity, based at least in part on the first activity data. The computing device receives a plurality of completed activity records related to a plurality of other users and trains an entity-identifying machine learning model to identify a plurality of entity-identifying characteristics related to a plurality of known entities to obtain a trained entity-identifying machine learning model. The computing device applies, when the predictive characteristic is present, the trained entity-identifying machine learning model to identify a known-entity data related to the completed activity of the user. The computing device updates the completed activity record of the completed activity of the user.