previous pauseresume next

NEWS

CmpE Has Multiple Open Positions. Deadline for application: 01 December 2020
  Bogazici University, Computer Engineering Department has multiple open Read more...
Boğaziçi University Leader in Artificial Intelligence, Big Data and Robotics
Tubitak has prepared a research ranking of Turkish universities in 131 subject Read more...
Collaboration between TUBITAK TUSSIDE and Industry 4.0 Platform
A new agreement between Boğaziçi University Industry 4.0 Platform and the Read more...
Boğaziçi University is ranked #1 among the Best Global Universities in Turkey
Boğaziçi University is ranked #1 among the Best Global Universities in Turkey, Read more...

CmpE Events

Yesterday

  1. CmpE Seminar: Onur Güngör ** Incorporating morphological information into neural named entity recognition taggers
    • Start time: 12:00pm, Tuesday, December 1st
    • End time: 01:00pm, Tuesday, December 1st
    • Where: Zoom
    • Title:  Incorporating morphological information into neural named entity recognition taggersSpeaker: Onur GüngörAbstract: Named entity recognition (NER) aims to extract mentions to several types of entities such as people, geographical names, and organizations in a given text. In this talk, I'll be presenting a neural named entity tagger model that exploits morphological information found in the surface form of words to improve the performance. The comparison between models that only use the word and character-based embeddings and models that add our morphological embeddings into the word representation shows that morphological analyses produced by morphological analyzers help to improve the NER performance, especially for morphologically rich languages. Our method does not require external morphological disambiguators, having morphological analyzers is enough. We also compare several ways to encode raw morphological analyses into morphological embeddings. This approach achieved state-of-the-art results in commonly used Turkish, Czech, and Hungarian NER datasets.About the speaker:Onur Güngör is a Ph.D. student at Boğaziçi University Computer Engineering Department. His research focuses on named entity recognition for morphologically rich languages, but he also writes papers about explaining NLP predictions, compiling interesting corpora, and correcting annoying spelling errors. He works as a senior data scientist at sahibinden.com where he develops systems that solve business problems using machine learning methods.

    • View this event in Google Calendar

Contact us

Department of Computer Engineering, Boğaziçi University,
34342 Bebek, Istanbul, Turkey

  • Phone: +90 212 359 45 23/24
  • Fax: +90 212 2872461
 

Connect with us

We're on Social Networks. Follow us & get in touch.