Svitlana Vakulenko

@svakulenk0

Postdoc researcher at ILPS, University of Amsterdam

Amsterdam, The Netherlands
Vrijeme pridruživanja: studeni 2012.

Tweetovi

Blokirali ste korisnika/cu @svakulenk0

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @svakulenk0

  1. Prikvačeni tweet
    20. stu 2019.
    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    26. sij

    With less than a week to go until our meetup with and , head over to and sign up!

    Prikaži ovu nit
    Poništi
  3. 17. sij

    full receptive field of self-attention + adaptive parallel computation via dynamic halting for Universal Transformer recurrent in depth

    Poništi
  4. 17. sij

    data augmentation by targeted sampling adding context for rare in distribution & ambiguous difficult words sorted by prediction loss for neural machine translation

    Poništi
  5. 17. sij

    KB-seq2seq embedding local KG constructed to compress relevant web documents for QA on ELI5 dataset with long answers and why/how questions from subreddit using indexed Common Crawl

    Poništi
  6. proslijedio/la je Tweet
    8. sij

    I am looking for postdoc interested in working on graph representation learning and graph neural networks! The start date and duration are flexible. If you are interested please submit your CV and a short cover letter to wlh@cs.mcgill.ca with "[POSTDOC APP]" in the subject line.

    Poništi
  7. proslijedio/la je Tweet
    6. sij

    Another internship opportunity at Microsoft Research Montreal in a project at the intersection of Data Protection/Privacy/FATE and Machine Learning. You'll have the chance to work with researchers from both the FATE and ML groups at MSR Montreal:

    Poništi
  8. proslijedio/la je Tweet
    8. sij

    Excited to announce our first SEA of the year, hosting two amazing speakers: Alexandra Olteanu () from Microsoft Research, New York and Svitlana Vakulenko () from ILPS, Amsterdam. Reserve your spot here:

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    30. pro 2019.

    Happy to announce that with a special focus on will be held in August 24–27 in Moscow, Russia! Have an interesting IR/ML/ConvAI topic to teach? Send a course proposals by February 29. More details:

    Poništi
  10. proslijedio/la je Tweet
    20. pro 2019.

    We have a fully funded PhD position in Knowledge Graphs for Conversational AI RT please

    Poništi
  11. proslijedio/la je Tweet
    12. pro 2019.

    The Amsterdam-based knowledge institutes are set to spend one billion euro on AI over the next ten years. Come and join the Amsterdam AI ecosystem!

    Poništi
  12. proslijedio/la je Tweet
    13. pro 2019.

    is about to present our work at workshop — stop by and try to challenge him ;-)

    Poništi
  13. 3. pro 2019.

    equity of attention with relevance constraint for fairness in ranking via ILP mediator accounts fragment&average user profiles into coherent chunks to trade-off utility for privacy in personalized search & recommendation

    Poništi
  14. proslijedio/la je Tweet

    . is solving unique challenges: Recommending the most relevant content out of billions of options in real time at scale. Today we’re sharing the first detailed overview of the elements that make Explore work:

    Poništi
  15. proslijedio/la je Tweet

    We have a vacancy for an asst. prof. on Hybrid Intelligence: AI systems that collaborate with people instead of replacing them. We offer a top-100 Dept (Times Higher), a top-20 city (Mercer) and a fully funded PhD student to supervise (closes 6 Jan)

    Poništi
  16. 30. stu 2019.

    ⚡️ unique opporunity to meet local IR & experts presenting their most recent work from top conferences! 5/5 on my evaluation metric: relevance+novelty+diversity+beautiful venue+stylish catering! Dank je

    Poništi
  17. proslijedio/la je Tweet

    If you are at , you are welcome to visit our posters on pre-trained language models for Dutch (ULMFiT and BERT-NL) Work with , , , , Data and models:

    Poništi
  18. proslijedio/la je Tweet

    Getting ready for post-DIR dinner 🍽️ – mjesto: Lloyd Hotel

    Poništi
  19. 29. stu 2019.

    choosing algorithms in focus groups study protyping user control mechanisms for transparency in personalised content by FairNews project

    Poništi
  20. proslijedio/la je Tweet
    29. stu 2019.

    Bias warning, but so far I'm really enjoying these shortform lightning talks . The authors are doing a great job at compressing complex papers in 10min talks!

    Poništi
  21. proslijedio/la je Tweet
    29. stu 2019.

    Two posters about and next to each other at . Great discussions during poster session.

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·