BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Date iCal//NONSGML kigkonsult.se iCalcreator 2.20.2//
METHOD:PUBLISH
X-WR-CALNAME;VALUE=TEXT:Eventi DIAG
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:STANDARD
DTSTART:20241027T030000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20240331T020000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RDATE:20250330T020000
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:calendar.28598.field_data.0@www.u-gov-ricerca.uniroma1.it
DTSTAMP:20260405T115704Z
CREATED:20241004T084450Z
DESCRIPTION:  AbstractEntity matching\, a core data integration problem\, i
 s the task of deciding whether two data tuples refer to the same real-worl
 d entity. Recent advances in deep learning methods\, using pre-trained lan
 guage models\, were proposed for resolving entity matching. Although demon
 strating unprecedented results\, these solutions suffer from a major drawb
 ack as they require large amounts of labeled data for training\, and\, as 
 such\, are inadequate to be applied to low resource entity matching proble
 ms. To overcome the challenge of obtaining sufficient labeled data we offe
 r a new active learning approach\, focusing on a selection mechanism that 
 exploits unique properties of entity matching. We argue that a distributed
  representation of a tuple pair indicates its informativeness when conside
 red among other pairs. This is used consequently in our approach that iter
 atively utilizes space-aware considerations. Bringing it all together\, we
  treat the low resource entity matching problem as a Battleship game\, hun
 ting indicative samples\, focusing on positive ones\, through awareness of
  the latent space along with careful planning of next sampling iterations.
  An extensive experimental analysis shows that the proposed algorithm outp
 erforms state-of-the-art active learning solutions to low resource entity 
 matching\, and although using less samples\, can be as successful as state
 -of-the-art fully trained known algorithms. BiographyAvigdor Gal is the Be
 njamin and Florence Free Chaired Professor of Data Science and the Co-chai
 r of the Center for Humanities & AI at the Technion - Israel Institute of 
 Technology. He is with the Faculty of Data & Decision Sciences\, where he 
 led the design of the first engineering program in data science in Israel 
 (and possibly the world). Gal’s research focuses on elements of data integ
 ration and process management and mining under uncertainty\, making use of
  state-of-the-art machine learning and deep learning techniques to offer a
 n improved data quality with about 150 publications in leading journals\, 
 books\, and conference proceedings (including multiple best paper and test
 -of-time awards). His research is implemented\, through his ties as a cons
 ultant\, in multiple industries including FinTech (e.g.\, Pagaya). In rece
 nt years\, with the increasing penetration of AI to all aspects of life\, 
 Gal has been involved in developing methods for embedding responsible AI i
 n companies and government authorities through an education process that i
 ncreases dialogue abilities between data scientists and other stake-holder
 s (e.g.\, lawyers and regulators).  
DTSTART;TZID=Europe/Paris:20241021T150000
DTEND;TZID=Europe/Paris:20241021T150000
LAST-MODIFIED:20241004T101906Z
LOCATION:Aula Magna\, DIAG
SUMMARY:The Battleship Approach to the Low Resource Entity Matching Problem
  - Avigdor Gal
URL;TYPE=URI:http://www.u-gov-ricerca.uniroma1.it/node/28598
END:VEVENT
END:VCALENDAR