Context-Aware Name Generation for Text Anonymization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for generating names are often generic and not customized for the entity being named, leading to uncontextualized pseudonyms or anonymization that can make text difficult to read.
Innovation Solution
A computing system utilizing machine-learned models to generate contextually appropriate names by receiving context data that describes entities to be named, and producing output data that includes names tailored to the described entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If generic name generation methods are used (random selection from predetermined list), then the system complexity is low and ease of manufacture is high, but the adaptability and customization for specific entities deteriorates
Solution Approach 1:
The patent transforms the name generation process from random selection to context-aware generation by changing the input parameters. The machine learning model receives context data about entities (attributes, relationships, domain information) and generates names tailored to specific entities rather than random generic names, directly resolving the contradiction between customization and complexity.
Solution Approach 2:
The patent replaces the simple mechanical approach of random selection from a predetermined list with an intelligent system using machine learning models. The ML model analyzes context data and generates appropriate names, substituting the crude mechanical method with a sophisticated computational approach that achieves customization without excessive complexity.
2Reliability
If simple redaction is used for anonymization, then the ease of operation is high and device complexity is low, but the readability and quality of the resulting text deteriorates
Solution Approach 1:
The patent changes the anonymization approach from simple redaction to intelligent name generation. Instead of merely removing or replacing names with generic placeholders, the system generates contextually appropriate pseudonyms that maintain the narrative flow and readability of the text, thereby improving reliability of anonymized output.
Solution Approach 2:
The machine learning model acts as an intermediary between the original text and the anonymized output. Rather than directly redacting names, the model generates intermediate pseudonyms that preserve the contextual meaning and readability while achieving anonymization, thus resolving the contradiction between simplicity and quality.
3Adaptability or versatility
If context-aware name generation using machine learning is implemented, then the adaptability and name relevance improve, but the computational resources and training requirements increase
Solution Approach 1:
The patent applies preliminary action by training the machine learning model in advance on large datasets containing entity information and corresponding names. This pre-training allows the model to learn patterns and generate contextually appropriate names during inference without requiring heavy computational resources at runtime, thus resolving the contradiction between adaptability and energy consumption.
Data Source
AI summary
A computing system can include one or more machine-learned models configured to receive context data that describes one or more entities to be named. In response to receipt of the context data, the machine-learned model(s) can generate output data that describes one or more names for the entity or entities described by the context data. The computing system can be configured to perform operations including inputting the context data into the machine-learned model(s). The operations can include receiving, as an output of the machine-learned model(s), the output data that describes the name(s) for the entity or entities described by the context data. The operations can include storing at least one name described by the output data.