Can computers detect sarcasm? Predict betrayal? Spot a drunk text?
Yes, all that and then some. A recent conference gave computational linguists a chance to show how sharp the cutting edge of language technology really is.
Linguist Daniel Midgley reads the articles so you don’t have to, for this episode of Talk the Talk.
Listen to this episode
Promo with Jane Hebiton
Coding in the classroom: Python overtakes French as most popular ‘language’ in primary schools
WaPo: Inside the surprisingly high-stakes quest to design a computer program that ‘gets’ sarcasm online
The Sarcasm Detector
The paper: Harnessing Context Incongruity for Sarcasm Detection (PDF)
ScienceNews: A few key signs betray betrayal
Cornell Uni: Language analysis predicts a coming betrayal
STUDY: Changes in language, tone could reveal impending betrayal
The paper: Linguistic Harbingers of Betrayal: A Case Study on an Online Strategy Game (PDF)
IIT-Bombay team creates program to detect drunk text message writers http://indianexpress.com/article/cities/mumbai/iit-bombay-team-creates-program-to-detect-drunk-text-message-writers/
The paper: A Computational Approach to Automatic Prediction of Drunk-Texting (PDF)
Paper: Modeling Argument Strength in Student Essays (PDF)
Paper: Automatic disambiguation of English puns (PDF)
Paper: Automatic Discrimination between Cognates and Borrowings (PDF)
All the ACL2015 papers!
The Gayby Project
Gayby Baby: NSW Government’s ban on screening same-sex parenting documentary sends ‘negative message’ to young people
Why banning Gayby Baby is damaging for gay families
Oh Hey There: Trisyllabic Lax
GLoWbE: The Corpus of Global Web-Based English
Find the tracks we play on the RTRFM webpage for this episode.
Image credit: http://www.aclweb.org/website_images/wRobot.png