Co-reference Resolution for Financial Transaction Intent Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current co-reference resolution techniques in natural language processing struggle to accurately link financial transaction events across multiple conversation threads, particularly in multiparty digital chat rooms, leading to missed opportunities and inefficiencies in financial trade negotiations.
Innovation Solution
A multi-step approach using a trained machine learning model for natural language processing to identify entities, classify intents, and link events within specific threads, enabling the validation and invalidation of intents by processing messages sequentially and maintaining a shared identifier for financial instruments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional co-reference resolution techniques are used to link named entities in conversations, then specific named entities can be linked together, but ambiguous intents and events cannot be accurately identified
Solution Approach 1:
The system segments the conversation analysis into multiple processing stages: entity identification, event detection, intent classification, and co-reference resolution. This multi-stage approach allows each component to specialize in specific tasks, improving overall accuracy while handling diverse financial transaction scenarios
Solution Approach 2:
The NLP system is designed to perform multiple functions simultaneously: it identifies named entities, detects financial events, classifies intents, and resolves co-references within a unified framework. This multi-functional approach enables the system to handle both specific entity linking and ambiguous intent identification with a single integrated solution
2Reliability
If all conversation messages are processed to identify financial transactions, then comprehensive transaction detection is achieved, but computational resources and processing time increase significantly
Solution Approach 1:
The system performs preliminary filtering and classification of conversation messages before detailed analysis. Messages are pre-processed to identify potential financial transaction indicators, allowing the system to focus computational resources only on relevant segments rather than processing entire conversation histories
Solution Approach 2:
The system extracts and isolates key financial transaction elements from conversational text, separating them from unrelated dialogue. This extraction approach allows the system to process only the essential transaction-related information, reducing overall processing time while maintaining detection completeness
3Measurement precision
If detailed analysis of each message is performed to accurately classify intents, then intent classification accuracy improves, but system complexity increases
Solution Approach 1:
The system applies different levels of analysis depth to different message types and conversation contexts. High-detail analysis is applied only where necessary for accurate intent classification, while simpler processing is used for routine or clearly-defined transaction types, optimizing the balance between accuracy and complexity
Data Source
AI summary
Systems and methods herein provide for understanding the context of multiple conversation events and accurately linking them together. Such may allow for fewer financial transaction opportunities to be missed and enable sell side institutions to book more trades. In one embodiment, a method of classifying financial transaction messages with a trained machine learning model includes identifying entities in a financial transaction message, identifying subsequent passages relating to the financial transaction message, and classifying intent as valid or invalid in the financial transaction message. The method also includes linking events within a specific thread by sequentially processing the passages of the financial transaction message.