kbp 2017 cold start
play

KBP 2017 Cold Start KB Construction and Slot Filling Hoa Dang - PowerPoint PPT Presentation

Dec 19, 2022 •320 likes •810 views

KBP 2017 Cold Start KB Construction and Slot Filling Hoa Dang Shahzad Rajput U.S. National Institute of Standards and Technology TAC 2017 Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 1 / 48 Outline Introduction 1 Task

  1. KBP 2017 Cold Start KB Construction and Slot Filling Hoa Dang Shahzad Rajput U.S. National Institute of Standards and Technology TAC 2017 Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 1 / 48

  2. Outline Introduction 1 Task Variants Changes in 2017 SF/KB Evaluation 2 Definitions Queries Participants Results Conclusion 3 Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 2 / 48

  3. Outline Introduction 1 Task Variants Changes in 2017 SF/KB Evaluation 2 Definitions Queries Participants Results Conclusion 3 Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 3 / 48

  4. Task Variants Knowledge Base Construction (K=Many) - Queries were not provided ED Evaluation Event Argument Evaluation Event Nugget Evaluation BeST Evaluation SF Evaluation (K=1,3) Slot Filling (K=1) - Queries involving SF slots were provided Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 4 / 48

  5. Task Variants Queries - Example # 1 LDC Query SF Query <query id=‘‘CS17_9999 ’’> <query id=‘‘ CSSF17_ENG_abcabdefde ’’> <mentions > <name >June McCarthy </name > <mention > <docid >ENG_142 </docid > <name >June McCarthy </name > <beg >16931 </beg > <docid >ENG_142 </docid > <end >16943 </end > <beg >16931 </beg > <enttype >PER </ enttype > <end >16943 </end > <slot >per:children </slot > </mention > <slot0 >per:children </slot0 > <mention > <slot1 >per:age </slot1 > <name >Junio McCarthy </name > </query > <docid >SPA_142 </docid > <query id=‘‘ CSSF17_SPA_defdeabcab ’’> <beg >2863 </beg > <name >Junio McCarthy <</name > <end >2869 </end > <docid >SPA_142 </docid > </mention > <beg >2863 </beg > </mentions > <end >2869 </end > <enttype >per </ enttype > <enttype >PER </ enttype > <nodeid >per_049 </ nodeid > <slot >per:children </slot > <slot0 >per:children </slot0 > <slot0 >per:children </slot0 > <slot1 >per:age </slot1 > <slot1 >per:age </slot1 > </query > </query > Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 5 / 48

  6. Task Variants Queries - Example # 2 LDC Query SF Query <query id=‘‘CS17_9999 ’’> <query id=‘‘ CSSF17_ENG_abcabdefde ’’> <mentions > <name >June McCarthy </name > <mention > <docid >ENG_142 </docid > <name >June McCarthy </name > <beg >16931 </beg > <docid >ENG_142 </docid > <end >16943 </end > <beg >16931 </beg > <enttype >PER </ enttype > <end >16943 </end > <slot >per:contact.meet_entity </slot > </mention > <slot0 >per:contact.meet_entity </slot0 > <mention > </query > <name >Junio McCarthy </name > <docid >SPA_142 </docid > <query id=‘‘ CSSF17_SPA_defdeabcab ’’> <beg >2863 </beg > <name >Junio McCarthy <</name > <end >2869 </end > <docid >SPA_142 </docid > </mention > <beg >2863 </beg > </mentions > <end >2869 </end > <enttype >per </ enttype > <enttype >PER </ enttype > <nodeid >per_049 </ nodeid > <slot >per:contact.meet_entity </slot > <slot0 >per:contact.meet_entity </slot0 > <slot0 >per:contact.meet_entity </slot0 > </query > </query > Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 6 / 48

  7. Task Variants Queries - Example # 3 LDC Query SF Query <query id=‘‘CS17_9999 ’’> <query id=‘‘ CSSF17_ENG_abcabdefde ’’> <mentions > <name >June McCarthy </name > <mention > <docid >ENG_142 </docid > <name >June McCarthy </name > <beg >16931 </beg > <docid >ENG_142 </docid > <end >16943 </end > <beg >16931 </beg > <enttype >PER </ enttype > <end >16943 </end > <slot >per:children </slot > </mention > <slot0 >per:children </slot0 > <mention > <slot1 >per:contact.meet_entity </slot1 > <name >Junio McCarthy </name > </query > <docid >SPA_142 </docid > <query id=‘‘ CSSF17_SPA_defdeabcab ’’> <beg >2863 </beg > <name >Junio McCarthy <</name > <end >2869 </end > <docid >SPA_142 </docid > </mention > <beg >2863 </beg > </mentions > <end >2869 </end > <enttype >per </ enttype > <enttype >PER </ enttype > <nodeid >per_049 </ nodeid > <slot >per:children </slot > <slot0 >per:children </slot0 > <slot0 >per:children </slot0 > <slot1 >per:contact.meet_entity </slot1 > <slot1 >per:contact.meet_entity </slot1 > </query > </query > Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 7 / 48

  8. Task Variants Queries - Example # 4 LDC Query SF Query <query id=‘‘CS17_9999 ’’> <query id=‘‘ CSSF17_ENG_abcabdefde ’’> <mentions > <name >June McCarthy </name > <mention > <docid >ENG_142 </docid > <name >June McCarthy </name > <beg >16931 </beg > <docid >ENG_142 </docid > <end >16943 </end > <beg >16931 </beg > <enttype >PER </ enttype > <end >16943 </end > <slot >per:is_liked_by </slot > </mention > <slot0 >per:is_liked_by </slot0 > <mention > <slot1 >per:age </slot1 > <name >Junio McCarthy </name > </query > <docid >SPA_142 </docid > <query id=‘‘ CSSF17_SPA_defdeabcab ’’> <beg >2863 </beg > <name >Junio McCarthy <</name > <end >2869 </end > <docid >SPA_142 </docid > </mention > <beg >2863 </beg > </mentions > <end >2869 </end > <enttype >per </ enttype > <enttype >PER </ enttype > <nodeid >per_049 </ nodeid > <slot >per:is_liked_by </slot > <slot0 >per:is_liked_by </slot0 > <slot0 >per:is_liked_by </slot0 > <slot1 >per:age </slot1 > <slot1 >per:age </slot1 > </query > </query > Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 8 / 48

  9. Task Variants Example#01: KB → EDL Knowledge Base :e4 type PER :e4 mention "Bart" Doc726 :37 -40 :e4 nominal_mention "brother" Doc726 :15 -21 EDL component EKB :e4_M1 Bart Doc726 :37 -40 NIL_0001 PER NAM 1.0 EKB :e4_M2 brother Doc726 :15 -21 NIL_0001 PER NOM 1.0 Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 9 / 48

  10. Task Variants Example#02: KB → EAL Knowledge Base :en1 type GPE :en1 mention " Springfield " Doc726 :402 -412 1.0 :ev1 type CONFLICT.ATTACK :ev1 mention.actual "bombing" Doc726 :418 -424 1.0 :ev1 conflict.attack:place.actual :en1 Doc726 :261 -431; Doc726 :402 -407; NIL 1.0 ... EAL component (arguments) EKB:ev1 .1 Doc726 Conflict.Attack Place Springfield 402 -407 261 -431 402 -407 NIL Actual 1.0 ... EAL component (linking) EKB:ev1.Doc726 EKB:ev1 .1 ... EAL component (corpuslinking) EKB:ev1 EKB:ev1.Doc726 ... Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 10 / 48

  11. Task Variants Example#03: KB → EN Knowledge Base :ev1 type CONFLICT.ATTACK :ev1 mention.actual "bombing" Doc726 :418 -424 1.0 EN component # BeginOfDocument Doc726 Doc726 E1 418 ,424 bombing Conflict_Attack Actual # EndOfDocument Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 11 / 48

  12. Task Variants Example#04: KB → BeST Knowledge Base :en1 type GPE :en1 mention " Springfield " Doc726 :402 -412 1.0 :en4 type PER :en4 mention "Bart" Doc726 :37 -40 :en4 per:dislikes :en1 Doc726 :422 -427 1.0 BeST component <?xml version ="1.0" encoding ="UTF -8"? > <belief_sentiment_doc id=" Doc726"> <sentiment_annotations > <entities > <entity ere_id =": en1 .500001" offset ="402" length ="11" > <text >Springfield </text > <sentiments > <sentiment polarity =" neg" sarcasm ="no" confidence ="1.0" > <source ere_id =": en4 .500001" offset ="37" length ="4" >Bart </ source > </sentiment > </sentiments > </entity > </entities > </ sentiment_annotations > </ belief_sentiment_doc > Hoa Dang, Shahzad Rajput (NIST) KBP 2017 Cold Start TAC 2017 12 / 48

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Off to a Cold Start a few observations November 2013 Ralph Grishman New

Off to a Cold Start a few observations November 2013 Ralph Grishman New

cold start NYU Off to a Cold Start a few observations November 2013 Ralph Grishman New York University Cold Start &gt; Slot Filling + EnEty Linking

156 views • 15 slides
NYU at Cold Start 2015: Experiments on KBC with NLP Novices Yifan He Ralph Grishman

NYU at Cold Start 2015: Experiments on KBC with NLP Novices Yifan He Ralph Grishman

NYU at Cold Start 2015: Experiments on KBC with NLP Novices Yifan He Ralph Grishman Computer Science Department New York University The KBP Cold Start Task and Common Approaches The KBP Cold Start task builds a knowledge base from

668 views • 33 slides
Item Cold-start Recommendations: Learning Local Collective Embeddings Martin Saveski MIT Media

Item Cold-start Recommendations: Learning Local Collective Embeddings Martin Saveski MIT Media

Item Cold-start Recommendations: Learning Local Collective Embeddings Martin Saveski MIT Media Lab Amin Mantrach Yahoo Labs Barcelona ACM Conference on Recommender Systems : 7 Oct 2014 Cold-Start When new user/item enters the system No

839 views • 35 slides
Cold Start 2016 Hoa Dang Shahzad Rajput National Institute of Standards and Technology TAC 2016

Cold Start 2016 Hoa Dang Shahzad Rajput National Institute of Standards and Technology TAC 2016

Cold Start 2016 Hoa Dang Shahzad Rajput National Institute of Standards and Technology TAC 2016 Hoa Dang, Shahzad Rajput (NIST) Cold Start 2016 TAC 2016 1 / 45 Outline Introduction 1 Task Variants Changes in 2016 KB Entity Discovery

1.18k views • 45 slides
Cold War Development of the Cold War The Cold War (1945-91) was one of perception where

Cold War Development of the Cold War The Cold War (1945-91) was one of perception where

Origins of the Cold War Development of the Cold War The Cold War (1945-91) was one of perception where neither side fully understood the intentions and ambitions of the other. This led to mistrust and military build-ups. United States

398 views • 37 slides
2017 COLD WAR RESEARCH PAPER AND PRESENTATION The Cold

2017 COLD WAR RESEARCH PAPER AND PRESENTATION The Cold

2017 COLD WAR RESEARCH PAPER AND PRESENTATION The Cold War is one of the three areas of focus and concentration for our IB History of

311 views • 5 slides
and the Cold Start Problem CS 278 | Stanford University | Michael Bernstein Last time We act

and the Cold Start Problem CS 278 | Stanford University | Michael Bernstein Last time We act

gettingstarted.md: Prototyping and the Cold Start Problem CS 278 | Stanford University | Michael Bernstein Last time We act differently in different spaces. Norms informal rules that govern behavior play a massive role in determining

1.48k views • 35 slides
Information Filtering for arXiv.org: Bandits, Exploration vs. Exploitation, and the Cold Start

Information Filtering for arXiv.org: Bandits, Exploration vs. Exploitation, and the Cold Start

Information Filtering for arXiv.org: Bandits, Exploration vs. Exploitation, and the Cold Start Problem Peter Frazier, Xiaoting Zhao School of Operations Research &amp; Information Engineering Cornell University Fusion Fest, DIMACS, Rutgers

254 views • 14 slides
Generation in Catalyst Bed for Solving of Cold Start Problem for Diesel Vehicles. Prof.

Generation in Catalyst Bed for Solving of Cold Start Problem for Diesel Vehicles. Prof.

Efficient Low Cost Technology for VOC Abatement in Off- Gases Based on Catalytic Oxidation With Ozone. Development of Catalytic Reactor Providing Direct Ozone Generation in Catalyst Bed for Solving of Cold Start Problem for Diesel

415 views • 19 slides
Cross-lingual Cold-Start Knowledge Base Construction M. Al-Badrashiny, J. Bolton5, A. T.

Cross-lingual Cold-Start Knowledge Base Construction M. Al-Badrashiny, J. Bolton5, A. T.

Cross-lingual Cold-Start Knowledge Base Construction M. Al-Badrashiny, J. Bolton5, A. T. Chaganty, K. Clark, C. Harman, L. Huang, M. Lamm, J. Lei, D. Lu, X. Pan, A. Paranjape, E. Pavlick, H. Peng, P. Qi, P. Rastogi, A. See, K. Sun, M.

1.01k views • 60 slides
Warm Up Cold-Start Advertisements Improving CTR predictions via Learning to Learn ID embeddings

Warm Up Cold-Start Advertisements Improving CTR predictions via Learning to Learn ID embeddings

Warm Up Cold-Start Advertisements Improving CTR predictions via Learning to Learn ID embeddings Feiyang Pan 1 , Shuokai Li 1 , Xiang Ao 1 , Pingzhong Tang 2 , Qing He 1 1 Institute of Computing Technology, CAS 2 Tsinghua University Feiyang

199 views • 16 slides
Transfer Learning from APP Domain to News Domain for Dual Cold-Start Recommendation Jixiong Liu 1

Transfer Learning from APP Domain to News Domain for Dual Cold-Start Recommendation Jixiong Liu 1

Transfer Learning from APP Domain to News Domain for Dual Cold-Start Recommendation Jixiong Liu 1 , Jiakun Shi 1 , Wanling Cai 1 , Bo Liu 2 , Weike Pan 1 Qiang Yang 2 and Zhong Ming 1 { 1455606137,1033150729,382970614 } @qq.com, {

475 views • 23 slides
Extreme cold start-up validation of wind turbine components by the use of a large climatic test

Extreme cold start-up validation of wind turbine components by the use of a large climatic test

Extreme cold start-up validation of wind turbine components by the use of a large climatic test chamber by pieterjan.jordaens@owi-lab.be www.owi-lab.be Understanding the topic: Why climate chamber testing? Wind turbines are installed word wide

758 views • 39 slides
Cold Start KB and Slot-Filling Approaches UMass Amherst Ben Roth, Nick Monath, David Belanger,

Cold Start KB and Slot-Filling Approaches UMass Amherst Ben Roth, Nick Monath, David Belanger,

Cold Start KB and Slot-Filling Approaches UMass Amherst Ben Roth, Nick Monath, David Belanger, Emma Strubell, Pat Verga and Andrew McCallum Outline Prediction Modules Universal Schema CNNs SVMs Rule-based Slot-Filling

557 views • 36 slides
Cold Brew THIS IS COLD BREW Cofgee brewed with cold fresh water over a long time gets unique

Cold Brew THIS IS COLD BREW Cofgee brewed with cold fresh water over a long time gets unique

Cold Brew THIS IS COLD BREW Cofgee brewed with cold fresh water over a long time gets unique fmavour characteristics. Cold Brew is already a hot trend in USA and Japan. Now we ofger the Swedish market an innovative product that matches an

75 views • 6 slides
Joe Ellis (presenter), Jeremy Getman, Stephanie Strassel Linguistic Data Consortium University of

Joe Ellis (presenter), Jeremy Getman, Stephanie Strassel Linguistic Data Consortium University of

Linguistic Resources for the 2015 TAC KBP Cold Start and Tri-Lingual Entity Discovery &amp; Linking Evaluations Joe Ellis (presenter), Jeremy Getman, Stephanie Strassel Linguistic Data Consortium University of Pennsylvania, USA Cold Start

495 views • 26 slides
What does a degree result mean? Charting UCLs journey from Honours Degree Classification to

What does a degree result mean? Charting UCLs journey from Honours Degree Classification to

What does a degree result mean? Charting UCLs journey from Honours Degree Classification to GPA History Honours Degree Classification created at Oxford in the early 19 th century when very few students, all from similar backgrounds and

604 views • 24 slides
Welcome! Boy Hennevelt, Marcel van Egmond, Toni van der Meer, Renata Longhi Sessions programme

Welcome! Boy Hennevelt, Marcel van Egmond, Toni van der Meer, Renata Longhi Sessions programme

Pre-Master February 2020 Welcome! Boy Hennevelt, Marcel van Egmond, Toni van der Meer, Renata Longhi Sessions programme Practical information The courses Milestones in Communication Science (12 ECTS) Introduction to Research

331 views • 17 slides
UNIVERSITY TECHNOLOGY STRATEGY Chris Manriquez, VP for Information Technology / CIO October 20,

UNIVERSITY TECHNOLOGY STRATEGY Chris Manriquez, VP for Information Technology / CIO October 20,

UNIVERSITY TECHNOLOGY STRATEGY Chris Manriquez, VP for Information Technology / CIO October 20, 2017 Agenda Global Context Strategic End State Where We Came From Digital Methods and Considerations Comprehensive Budget Plan /

510 views • 23 slides
Pointer Analysis: The Big Picture View Uday Khedker (www.cse.iitb.ac.in/uday) Department of

Pointer Analysis: The Big Picture View Uday Khedker (www.cse.iitb.ac.in/uday) Department of

Pointer Analysis: The Big Picture View Uday Khedker (www.cse.iitb.ac.in/uday) Department of Computer Science and Engineering, Indian Institute of Technology, Bombay Dec 2017 WSSE Pune PTA Big Picture: The Big Picture 1/22 Outline The

831 views • 67 slides
SIRENA SIRENA External EMC simulation for radio electric systems, in the close environment of

SIRENA SIRENA External EMC simulation for radio electric systems, in the close environment of

SIRENA SIRENA External EMC simulation for radio electric systems, in the close environment of the aircraft Aeronautics Days 2006, Vienna 19-21 June SIRENA technical main objectives technical main objectives SIRENA Development of software and

1.05k views • 32 slides
The Closeout State of the Faint Object Spectrograph Charles D. (Tony) Keyes Space Telescope

The Closeout State of the Faint Object Spectrograph Charles D. (Tony) Keyes Space Telescope

1997 HST Calibration Workshop Space Telescope Science Institute, 1997 S. Casertano, et al., eds. The Closeout State of the Faint Object Spectrograph Charles D. (Tony) Keyes Space Telescope Science Institute, 3700 San Martin Drive, Baltimore MD

324 views • 10 slides
Seasonal signals induced by monument thermal effects: Evidence in GPS position time series of

Seasonal signals induced by monument thermal effects: Evidence in GPS position time series of

Second Workshop of DAAD Thematic Network Seasonal signals induced by monument thermal effects: Evidence in GPS position time series of short baselines KH Wang 1,2 , WP Jiang 1 , H Chen 1 , XD An 1 , XH Zhou 1 , P Yuan 1,2 , QS Chen 1 1 GNSS

904 views • 36 slides
Matrix Metals Limited Presentation October 2008 Disclaimer The material used is this

Matrix Metals Limited Presentation October 2008 Disclaimer The material used is this

Matrix Metals Limited Presentation October 2008 Disclaimer The material used is this presentation is intended to be a summary of selected geological data, current and proposed activities, as well as resource estimates based on information

389 views • 26 slides

More recommend