Scientists at the University of Sharjah report that they have developed a new machine-learning system designed to overcome challenges encountered in the diacritization of Arabic texts. The system mainly targets problems that existing programs face when encountering undiacritized Arabic script, writing that lacks the vowel marks necessary to pronounce words correctly, a process linguists refer to as diacritization. The presence of diacritics in Arabic is vital not only for how a word is pronounced, but also for semantics. A single word can have multiple, entirely different meanings, depending on how it is articulated.

Early cancer detection often relies on complex, invasive, and time-consuming staining procedures. A research team from Southeast University has developed a novel "label-free" biosensor that uses the physics of "band folding" to unlock high-density hidden electromagnetic modes in the sub-terahertz range. This technology creates a rich spectral fingerprint that can distinguish between normal cells and cancer cells of varying malignancy without chemical markers. By correlating macroscopic electromagnetic signals with microscopic cellular biomass density, this method offers a safe, rapid, and non-destructive path for future clinical cancer screening.