Machine-Learned Annotator for Financial Transaction Disambiguation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Financial transaction data often contains ambiguous entities, making it difficult to determine the identity of involved parties or businesses, which hinders user queries and analysis.
Innovation Solution
A machine-learning based annotator is trained to disambiguate financial transaction data by identifying entities using labeled training data, including text and entity associations, and annotates the data with unique identifiers, allowing for accurate entity recognition and user queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If financial transaction data is stored with minimal information, then storage efficiency is improved, but entity disambiguation capability deteriorates
Solution Approach 1:
The system segments entity identification into multiple stages: initial data storage with transaction descriptors, followed by separate disambiguation processing that matches descriptors against entity databases. This allows efficient storage while enabling accurate entity identification when needed.
Solution Approach 2:
The system performs preliminary actions by pre-building entity databases with known entity information and training machine learning models before actual transaction disambiguation. This preparation enables faster and more accurate entity identification during query processing without increasing transaction data storage requirements.
2Measurement precision
If machine learning models are trained with extensive labeled data, then entity recognition accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The system uses partial action by training machine learning models on a representative subset of labeled transaction data rather than exhaustively processing all possible transactions. This approach achieves sufficient accuracy for disambiguation while significantly reducing training time and computational resources.
Solution Approach 2:
The system performs preliminary action by pre-training machine learning models offline using historical labeled data before deployment. This allows the models to learn entity patterns in advance, enabling fast real-time disambiguation during actual use without requiring extensive processing time during operational queries.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving data that represents actions taken by a particular user with respect to entities of a plurality of entity types, disambiguating the data by identifying entities specified in the data using a machine learning based annotator that is trained to recognize entities and annotating the data, receiving a query specific from the particular user, querying the data that represents actions taken by the particular user, generating an answer to the query from the data representing the actions taken by the particular user, and providing the answer to the query.