Predicting Custom Fields from Transaction Text
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in predicting custom field names from free form text in transaction data, as users enter unstructured information that may relate to specific custom fields, but existing technologies lack effective methods to automatically identify and suggest relevant field names and types.
Innovation Solution
A method and system that normalize transaction text using machine learning models to predict custom field names and types by analyzing historical transaction data, generating prediction data, and presenting these predictions to user devices for updating transaction data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If free form text is used for transaction descriptions, then users can enter flexible unstructured information, but it becomes difficult to automatically identify and predict relevant custom field names
Solution Approach 1:
The patent introduces an intermediary processing layer between free form text input and custom field identification. This layer includes text normalization components that convert unstructured text into a standardized format, and prediction components that analyze the normalized text to suggest relevant custom field names. This intermediary processing resolves the contradiction by maintaining input flexibility while enabling automated field detection.
Solution Approach 2:
The patent replaces manual mechanical processes of field identification with automated computational systems. Machine learning models and natural language processing algorithms automatically analyze transaction text and predict custom field names, eliminating the need for manual field identification and resolving the difficulty of detecting relevant fields from unstructured text.
2Measurement precision
If manual identification of custom fields is performed, then accuracy can be maintained, but time consumption and user effort increase significantly
Solution Approach 1:
The patent performs preliminary actions by pre-processing transaction text through normalization before field identification. The system normalizes text, extracts features, and generates predictions in advance, which reduces the time required for actual field identification while maintaining accuracy through multiple processing stages.
Solution Approach 2:
The patent implements feedback mechanisms where prediction results are presented to users for validation and correction. User feedback on prediction accuracy is used to refine and improve the prediction model over time, maintaining high accuracy while reducing time consumption through increasingly precise automated predictions.
3Measurement precision
If text normalization is applied to free form text, then prediction accuracy improves, but processing complexity increases
Solution Approach 1:
The patent segments the text normalization and prediction process into distinct modular components. Each component handles a specific aspect of text processing (normalization, feature extraction, prediction), which reduces overall processing complexity while maintaining prediction accuracy through specialized processing at each stage.
Data Source
AI summary
A method predicts custom fields from text. Transaction text is normalized from transaction data to generate normalized text. A field prediction and a type prediction are selected using prediction data and the normalized text. The prediction data is generated using a machine learning model trained to identify field predictions from free form text. The field prediction and the type prediction are presented to a client device. In response to user input from the client device, the transaction data is updated with the field prediction.