BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//bbuzz22//speaker//89XRTP
BEGIN:VTIMEZONE
TZID:Europe/Berlin
BEGIN:DAYLIGHT
DTSTART:20210613T000000
TZNAME:CEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20211031T030000
RDATE:20221030T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20220327T030000
RDATE:20230326T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Hybrid search > sum of its parts? - Lester Solbakken
DTSTART;TZID=Europe/Berlin:20220613T172000
DTEND;TZID=Europe/Berlin:20220613T180000
DTSTAMP:20261009T192855Z
UID:pretalx-bbuzz22-YEHRTE@pretalx.com
DESCRIPTION:Over the decades\, information retrieval has been dominated by
  classical methods such as BM25. These lexical models are simple and effec
 tive yet vulnerable to vocabulary mismatch. With the introduction of pre-t
 rained language models such as BERT and its relatives\, deep retrieval mod
 els have achieved superior performance with their strong ability to captur
 e semantic relationships. The downside is that training these deep models 
 is computationally expensive\, and suitable datasets are not always availa
 ble for fine-tuning toward the target domain.\n\nWhile deep retrieval mode
 ls work best on domains close to what they have been trained on\, lexical 
 models are comparatively robust across datasets and domains. This suggests
  that lexical and deep models can complement each other\, retrieving diffe
 rent sets of relevant results. But how can these results effectively be co
 mbined? And can we learn something from language models to learn new index
 ing methods? This talk will delve into both these approaches and exemplify
  when they work well and not so well. We will take a closer look at differ
 ent strategies to combine them to get the best of both\, even in zero-shot
  cases where we don't have enough data to fine-tune the deep model.\n\nThe
  Search track is presented by OpenSource Connections
LOCATION:Kesselhaus
URL:https://pretalx.com/bbuzz22/talk/YEHRTE/
END:VEVENT
END:VCALENDAR