Graham, J. (1999). The reader’s helper: A personalized document reading environment. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 481–488

https://doi.org/10.1145/302979.303139

An Augmented reading intended to help people read “horizontally”, rather than “vertically”—i.e. skimming across many documents to find relevant information (see Syntopic reading, Skillful reading is often non-linear). Key design elements include:

• a scoring model which helps readers decide which papers might be most relevant to their interests
• automatic highlighting of passages in a paper which might be most important, or relevant to reader interest
• a minimap which allows rapid navigation and orientation with respect to those highlights (or readers’ own)
• simple automatic summarization of a document as it relates to a specific quer

Similar core ideas recapitulated later, using modern transformer models, in Fok, R., Kambhamettu, H., Soldaini, L., Bragg, J., Lo, K., Hearst, M., Head, A., & Weld, D. S. (2023). Scim: Intelligent Skimming Support for Scientific Papers. Proceedings of the 28th International Conference on Intelligent User Interfaces, 476–490.

Curious to note that the author, Jamey Graham, now works at Apple (doing ML stuff…)