Native Language Prediction via Eye Gaze Fixation Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining a person's native language rely solely on written samples and do not utilize eye gaze or eye tracker technology effectively.
Innovation Solution
A method that uses eye tracker cameras to record gaze location during sentence reading, extracting linguistically motivated features from the raw eye gaze data, and employs a supervised machine learning approach, such as a neural network classifier, to predict the native language of the user.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If eye tracker technology is used to predict native language, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent uses gaze patterns as an intermediary between the eye tracker and native language identification. Instead of directly analyzing complex eye movements, the system extracts linguistic features from gaze data (such as fixation durations, saccade patterns, and regression behaviors) that serve as mediators to infer native language with high accuracy while managing system complexity.
Solution Approach 2:
The patent replaces traditional mechanical language assessment methods (written samples, interviews) with an optical-based eye tracking system. By substituting the mechanical/behavioral assessment with optical gaze measurement and computational analysis, the system achieves higher objectivity and precision in native language identification.
2Reliability
If gaze pattern analysis is used instead of written samples, then reliability is improved, but loss of information increases
Solution Approach 1:
The patent transitions from analyzing linguistic output (written samples) to analyzing oculomotor behavior (gaze patterns) as a different dimension of language processing. By measuring where and how long readers fixate on text, the system captures implicit linguistic knowledge without requiring explicit language production, thereby maintaining reliability while preserving linguistic information.
Solution Approach 2:
The eye tracking system allows the reading process itself to reveal native language information without requiring the reader to consciously produce language samples. The gaze patterns naturally self-reveal linguistic background through automatic oculomotor responses to linguistic features, eliminating the need for separate language production tasks.
3Productivity
If supervised machine learning is employed for language prediction, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent employs pre-trained machine learning models that have been trained on large datasets of gaze patterns and native language information. This preliminary training allows the system to quickly predict native language during actual use without performing complex training computations in real-time, thereby achieving high productivity while managing processing complexity through offline preparation.
Data Source
AI summary
In an embodiment, a method includes presenting, on a display, sample text in a given language to a user. The method further includes recording eye fixation times for each word of the sample text for the user and recording saccade times between each pair of fixations of the sample text. The method further includes comparing features of the gaze pattern of the user to features of a gaze pattern of a plurality of training readers. Each training reader (e.g., training user) has a known native language. The method further generates a probability of at least one an estimated native language of the user based on the results of the comparison.