This spring, five New York City graduate programs ran a course in partnership aimed at addressing the key challenges of our information ecosystem. How does the information ecosystem contribute to the health of representative democracies such as the...
Word embeddings are a popular machine-learning method that represents each English word by a vector, such that the geometry between these vectors captures semantic relations between the corresponding words. We demonstrate that word embeddings can be...
Ani NenkovaThe gender associations in embeddings closely track the proportion of that gender in the profession pnas.org/content/115/16…@jurafsky
So an effective way to change word representations would be to change society.
Ani NenkovaEMNLP-IJCNLP 2019 website is up! Submissions due May 21, 2019
Mark your calendars and make sure you fill in the SIGDAT survey about your preferences on location of EMNLP 2020 (Mexico, Puerto Rico, Ireland, UK, Eastern USA)
#nlp #NLProc #emnlp
Ani Nenkova@emilymbender Marie-Catherine de Marneffe, Marta Recasens, Christopher Potts:
Modeling the Lifespan of Discourse Entities with Application to Coreference Resolution. J. Artif. Intell. Res. 52: 445-475 (2015)