Bob West

@cervisiarius

Tenure-track assistant professor at EPFL, Data Science Lab (dlab)

Joined September 2010

Tweets

You blocked @cervisiarius

Are you sure you want to view these Tweets? Viewing Tweets won't unblock @cervisiarius

  1. Nov 23

    The whole T-shirt (or D-shirt, as I prefer to call it) is actually a verbatim quote by Tristan Tzara about the Dada movement, but with "data" plugged in for "dada". :)

    Undo
  2. Retweeted
    Nov 23

    Bob West () has openings for students in his group. Find out more about his at: , and learn more about our PhD program:

    Undo
  3. Retweeted
    Nov 16

    Our paper on the impact of the COVID-19 crisis on Wikipedia has just been accepted at ICWSM-21 🎉 This is joint work with (from 's DLAB) + & (from Aachen). 📜: 🧵/

    Show this thread
    Undo
  4. Nov 13

    About 2 weeks to the tutorial proposal deadline for ! (Deadline: 30 Nov) There's a specific category for hands-on tutorials. Being hands-on is difficult in these online-only times (the conf will be online only...), but thus all the more important!

    Undo
  5. Nov 3

    As a co-chair (w/ ) for hands-on tutorials at , I'm sharing the call for tutorials: NEW: there will be 2 kinds of tutorials: (1) lecture-style, (2) hands-on. Please share the call, and consider submitting a proposal yourself!

    Undo
  6. Retweeted
    Oct 23

    Nicolas Flammarion is looking for student. Find out more about his , and learn more about our PhD program:

    Undo
  7. Retweeted
    Oct 29

    Rachid Guerraoui is looking for students! Learn about his and find out more about our PhD program:

    Undo
  8. Retweeted
    Oct 22

    When toxic web communities get banned, they don’t disappear... Rather, they migrate to a new platform! In that context, it is worth asking: -> Should platforms ban these communities? Our new pre-print tackles this question 📜: 1/

    , , and 2 others
    Show this thread
    Undo
  9. Retweeted
    Oct 14

    Wenzel Jakob () is looking for students. Find out more about his , and learn more about our PhD program:

    Undo
  10. Retweeted
    Oct 1

    "Crosslingual Topic Modeling with WikiPDA" Wikipedia-based Polyglot Dirichlet Allocation, a crosslingual topic model that represents Wikipedia articles written in any language as distributions over language-independent topics. (Piccardi and West, 2020)

    , , and 3 others
    Undo
  11. Sep 9

    We hope G-TAB will boost the potential of Google Trends as a source of insights! And if nothing else, we’re adding to the literature one more case where e is the optimal constant 😀 A big yeehah to , who implemented the Python API in his internship!

    Show this thread
    Undo
  12. Sep 9

    G-TAB solves these problems. It expresses the popularity of an arbitrary number of Google queries on a common scale without being compromised by rounding errors. How? By combining careful offline preprocessing with efficient online search.

    Show this thread
    Undo
  13. Sep 9

    Google Trends is a Swiss army knife for data scientists & journalists, pioneered by , , et al. But GT data can be tough to handle: only up to 5 queries can be compared at a time; results are relative & rounded to integers, such that often no useful info is left:

    Show this thread
    Undo
  14. Sep 9

    Thrilled to release Google Trends Anchor Bank (G-TAB), a tool that will enable data scientists to use Google Trends in ways that were not possible before. More details in the thread below!

    , , and 6 others
    Show this thread
    Undo
  15. Retweeted
    Sep 8

    📢 Our call for workshops at is now open! Find out more here: ⚠️4 Days left to submit a for ! Details for submission:

    Undo
  16. Sep 4

    So far, Wikipedia's full revision history has been available only in wiki markup, not in HTML -- a big limitation for researchers. We are changing this by releasing WikiHist.html, Wikipedia's full history (up until March 2019) in HTML: Caveat emptor: 7 TB!

    Undo
  17. Aug 21

    In times of the pandemic, we're experimenting with a new format: instead of one big monolithic AMLD event, we'll have a series of smaller installments through the winter/spring: one topically coherent track per month. Consider organizing one of these tracks! Details below!

    Undo
  18. Retweeted
    Aug 20

    Our pre-print “Experts and Authorities receive disproportionate attention on Twitter during the COVID-19 crisis” is out! 📜 Amazing work w Follow the 🧵 for findings 1/

    Show this thread
    Undo
  19. Retweeted
    Jul 15

    New dataset out: "Wikipedia Citations, a comprehensive dataset of citations with identifiers extracted from English Wikipedia" Pre-print Dataset Code

    Show this thread
    Undo
  20. Retweeted
    May 20

    "WikiHist.html: English Wikipedia’s Full Revision History in HTML Format", showing that over half of the wiki links present in HTML are missing from raw wikitext. (Mitrevski et al, 2020) paper: data and code:

    Undo

Loading seems to be taking a while.

Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

    You may also like

    ·