Improving text comprehensibility through Large Language Models
Tools that simplify and make technical or complex texts more accessible in a semi-automatic way, using advanced language models (Large Language Models) followed by human review.
Dedicated indicators allow users to measure the desired level of readability — for example, making texts easily understandable for those less familiar with complex content — evaluate the improvements made, and ensure that the original meaning is accurately preserved.
Ability to compare the modified text with the original and make revisions
Use of standard metrics and setting of thresholds to be achieved
Interventions on lexical paraphrasing, syntax reordering, and content reorganization
Combination of human expertise and automation for scalability and timeliness
Ability of the model to explain the operations performed (explainability)
Use case
Summaries of resolutions, acts, and documents intended for citizens
Support for public communication on institutional portals and channels
Simplification of manuals and procedures, including for internal operators
