Abstract
Systems for contextual text analysis can provide results based on a user's selection but may lack transparency in their reasoning. This can be a challenge when a system misinterprets the semantic context of an ambiguous word or phrase. A described technique can address this by using a machine learning model to analyze a user's initial text selection along with its surrounding text. The model can predict an expanded, semantically coherent conceptual unit that may represent a user's likely intent. This system-generated expansion can be displayed to the user, for example, through a multi-tiered visual rendering that distinguishes it from the original selection. This approach can provide users with insight into the system's contextual analysis, enabling them to verify or correct the selection and thereby provide a potentially more accurate input for downstream operations.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Miglino, Jerome; Hickey, Brendan; Leszczuk, Will; and Shi, MJ, "Model-Based Semantic Expansion of Text Selections with Multi-Tiered Visual Rendering", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9743