Bilkent University
Department of Computer Engineering


Speak Memory: Gender Distinction and Language Change Quantification from Memoirs of Old-time Movie Goers


Sevil Çalışkan
MS Student
Computer Engineering Department
Bilkent University

Discourse analysis receive considerable attention with the internet era as more and more people are connected, and more tools become available to analyze. Discourse analysis, or analysis of spoken language, is conducted by statistical methods with NLP and stylometric tools. Understanding discourse and its variation is important both in socio linguistic studies and intelligence applications. In this work, we study on the spoken memoirs of a group of old-time movie goers of different age groups, as they tell their experiences by answering a set of questions. Several experiments are performed to see the gender distinction of the speakers and language change among age groups. Results show that gender can be predicted perfectly as observed in written texts. Our analysis on language change among other aspects consider token and type length change with age.


DATE: 29 April 2019, Monday, CS590 presentations begin at @ 15:40