BEGIN:VCALENDAR
CALSCALE:GREGORIAN
VERSION:2.0
METHOD:PUBLISH
PRODID:-//Drupal iCal API//EN
X-WR-TIMEZONE:America/New_York
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
DTSTART:20070311T020000
TZNAME:EDT
TZOFFSETTO:-0400
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
DTSTART:20071104T020000
TZNAME:EST
TZOFFSETTO:-0500
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
SEQUENCE:1
X-APPLE-TRAVEL-ADVISORY-BEHAVIOR:AUTOMATIC
234721
20260414T095946Z
DTSTART;TZID=America/New_York:20260421T143000
DTEND;TZID=America/New_York:2
 0260421T153000
URL;TYPE=URI:https://www.wpi.edu/news/calendar/events/compu
 ter-science-department-ms-thesis-presentation-nicholas-pulsone-budget-awar
 e-entity-matching
Computer Science Department, MS Thesis Presentation, Nicholas Pulsone  " Budget-Aware Entity Matching Across Domains"
Nicholas Pulsone\nMS StudentWPI – Computer Science Department\nTuesday, April 21, 2026\nTime:
  2:30 p.m. – 4:00 p.m.Location: Fuller Labs 141\nZoom Link: https://wpi.
 zoom.us/j/97178831714\nCommittee members :Advisor: Prof. Roee ShragaReader
 :  Prof. Fabricio Murai\nAbstract:\nEntity Matching (EM)--the task of dete
 rmining whether two data records refer to the same real-world entity--is a
  core task in data integration. Recent advances in deep learning have set 
 a new standard for EM, particularly through fine-tuning Pretrained Languag
 e Models (PLMs) and, more recently, Large Language Models (LLMs). However,
  fine-tuning typically requires large amounts of labeled data, which are e
 xpensive and time-consuming to obtain.In the context of e-commerce matchin
 g, labeling scarcity varies widely across domains, raising the question of
  how to intelligently train accurate domain-specific EM models with limite
 d labeled data. In this work we assume users have only a limited amount of
  labels for a specific target domain but have access to labeled data from 
 other domains.We introduce BEACON, a distribution-aware, budget-aware fram
 ework for low-resource EM across domains. BEACON leverages the insight tha
 t embedding representations of pairwise candidate matches can guide the ef
 fective selection of out-of-domain samples under limited in-domain supervi
 sion. We conduct extensive experiments across multiple domain-partitioned 
 datasets derived from established EM benchmarks, demonstrating that BEACON
  consistently outperforms state-of-the-art methods under different trainin
 g budgets.\n
END:VEVENT
END:VCALENDAR