Karaoglan, BaharAslan, SinemKarayer, ErdemAshyralyev, ALukashov, A2019-10-272019-10-272012978-0-7354-1077-00094-243X0094-243Xhttps://doi.org/10.1063/1.4747687https://hdl.handle.net/11454/454581st International Conference on Analysis and Applied Mathematics (ICAAM) -- OCT 18-21, 2012 -- Gumushane, TURKEYIn this study we show empirically on a Turkish corpus that Zipf's power laws relating frequencies of words to rank, vocabulary and number of meanings can also be derived from term x document analysis. This supports the idea put forward by Ding that Zipfian distributions and SVD carry similar hidden (latent) semantics.en10.1063/1.4747687info:eu-repo/semantics/closedAccessZipfian distributionsterm document matrixterm similaritiesConference Object1470247250WOS:000309524400063N/AN/A