Skip to content

Runs - Web 2014

CiirAll1

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: CiirAll1
  • Participant: UMASS_CIIR
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 3953061d1887ff3299308a53d05ed404
  • Run description: combination of 40 features, entity context features using knowledge base links and baseline scores using Galago and Indri with FACC1 annotations reranking with RankLib, coordinate ascent, optimize for ERR@20, no normalization, 1 start for details see Dalton, Dietz, Allan. SIGIR 2014. uses wikipedia wex dump from january 2012

CiirSdm

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: CiirSdm
  • Participant: UMASS_CIIR
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: adhoc
  • MD5: 53b35674699d4272d49ad75d4a875386
  • Run description: Indri sequential dependence model with standard parameters 0.8, 0.15, 0.05

CiirSub1

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: CiirSub1
  • Participant: UMASS_CIIR
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: d2ddb68edbc9b273e9e80e5f63d81d93
  • Run description: CiirSub1 combination of 13 features, entity context features using knowledge base links and sdm scores using Galago and Indri with FACC1 annotations reranking with RankLib, coordinate ascent, optimize for ERR@20, no normalization, 1 start for details see Dalton, Dietz, Allan. SIGIR 2014. uses wikipedia wex dump from january 2012 features: sdm, wikirm, collection-psg, contextFeatsentity-8, contextFeats-idQL-enti ty-50-20, names-mention-numEnts20, wikipedia-20, contextFeats-names-descentity-50, linkedEnts-top1-idQl-20, top1names-numEnts10, wiki-idQl-50, wiki-categoryQl-1, wiki -typeQl-5

CiirSub2

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: CiirSub2
  • Participant: UMASS_CIIR
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 78421cc08c7b096e47864d82a7846c0b
  • Run description: CiirSub2 combination of 13 features, entity context features using knowledge base links and sdm scores using Galago and Indri with FACC1 annotations reranking with RankLib, coordinate ascent, optimize for ERR@20, no normalization, different start seed for details see Dalton, Dietz, Allan. SIGIR 2014. uses wikipedia wex dump from january 2012 features: sdm, wikirm, collection-psg, contextFeatsentity-8, contextFeats-idQL-entity-50-20, names-mention-numEnts20, wikipedia-20, contextFeats-names-descentity-50, linkedEnts-top1-idQl-20, top1names-numEnts10, wiki-idQl-50, wiki-categoryQl-1, wiki-typeQl-5

CiirWikiRm

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: CiirWikiRm
  • Participant: UMASS_CIIR
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: adhoc
  • MD5: 10bd31dd1219208814d70ab74a5c8de2
  • Run description: SDM with wikipedia expansion model (generated with Indri) SDM default parameters 0.8, 0.15, 0.05 RM weight 0.8/0.2 uses wikipedia wex dump from january 2012

ICTNET14ADR1

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: ICTNET14ADR1
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/9/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 53922962410acc54bf4a63067bd50d95
  • Run description: The wikipedia data along with ClueWeb12 are used to identify the phrase in the query, and the redirect of the phrase are got. BM25 are applied to anchor, title and body fields. GibbsLDA++ is used to get subtopics and than greedy select method is used to generate the final RUN.

ICTNET14ADR2

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: ICTNET14ADR2
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/9/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 40bce7b7262b7cdf675bc8e2b6b2090e
  • Run description: Same as ICTNET14ADR1, only the greedy select methods are different.

ICTNET14ADR3

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: ICTNET14ADR3
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/8/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 36210b657796552de73d14d53e5f45b9
  • Run description: apply BM25 on anchor and title field, and proximity BM25 on body field.

ICTNET14RSR1

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: ICTNET14RSR1
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/10/2014
  • Type: automatic
  • Task: risk
  • MD5: b7f8ef70e9b72aa05030ef8e67a1394c
  • Run description: apply fusion method on official Indri baseline and ICTNET14ADR1

ICTNET14RSR2

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: ICTNET14RSR2
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/10/2014
  • Type: automatic
  • Task: risk
  • MD5: 64d6059bb4c5cec0811fe30966689338
  • Run description: apply fusion method on official Terrier baseline and ICTNET14ADR1

ICTNET14RSR3

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: ICTNET14RSR3
  • Participant: ICTNET
  • Track: Web
  • Year: 2014
  • Submission: 8/10/2014
  • Type: automatic
  • Task: risk
  • MD5: 59f66581fa3fbbbc248afa1515218e68
  • Run description: apply fusion method on ICTNET14ADR1 and ICTNET14ADR3

Protoss

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: Protoss
  • Participant: Group.Xu
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 5ba18cefc9bed51d0f9cfb5af6cd5c5b
  • Run description: LETOR and basic QE is used base on the primitive run. LETOR adopts Listwise Genetic Programming, but risking of overfitting

SNUMedinfo11

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: SNUMedinfo11
  • Participant: SNUMedinfo
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 6e77a4428873b705a9c476a15a707f51
  • Run description: Sequential Dependence model

SNUMedinfo12

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: SNUMedinfo12
  • Participant: SNUMedinfo
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 82b93ce0fd7fbdcf0337ecb1bfb89b63
  • Run description: Sequential Dependence model + Waterloo spam filtering

SNUMedinfo13

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: SNUMedinfo13
  • Participant: SNUMedinfo
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 4cebd3edf9ee4f314dbb93c69666c4d9
  • Run description: Sequential Dependence model + Waterloo spam filtering + Pagerank Prior

Terran

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: Terran
  • Participant: Group.Xu
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: fb5f5202a07439084909618bbc8728c6
  • Run description: LETOR and basic QE is used base on the primitive run.

udel_itu

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: udel_itu
  • Participant: udel
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 85377fdb02b23be9806065949fee6439
  • Run description: indri using evidence from inlink text, titles, and urls only (no body text).

udel_itub

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: udel_itub
  • Participant: udel
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 41ac92e0eb73377f6b3cb2c799c83afd
  • Run description: indri using evidence from inlink text, titles, urls, and body text.

udelCombCAT2

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: udelCombCAT2
  • Participant: udel
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: risk
  • MD5: 6d18cac302fbbc47578c6b1721226e36
  • Run description: Use Bing to get query suggestions for each query. Submit the query to Indri. Then aggregate the results using CombCAT (a variant of CombMNZ). Essentially a diversity run for which we investigate the robustness.

UDInfoWebAX

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: UDInfoWebAX
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 1c1df7cb1f7d6583c8b4f0cde4eda6fc
  • Run description: Axiomatic approach

UDInfoWebENT

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: UDInfoWebENT
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: adhoc
  • MD5: 619c8ad5011f702f26787ef9cb17ed21
  • Run description: Entity Centric Query Expansion

UDInfoWebLES

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: UDInfoWebLES
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: adhoc
  • MD5: 8c02fd5a082e06403227427d83d0ea3b
  • Run description: Latent Entity Space

UDInfoWebRiskAX

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: UDInfoWebRiskAX
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: risk
  • MD5: 3d94110515eb93d4643ae44678474d92
  • Run description: Optimized based on UDInfoWebLES.

UDInfoWebRiskRM

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: UDInfoWebRiskRM
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: risk
  • MD5: ce6b53d147136416e5d96d4e4b95e7f2
  • Run description: Optimized based on the official Indri 2014 baseline run.

UDInfoWebRiskTR

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: UDInfoWebRiskTR
  • Participant: udel_fang
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: manual
  • Task: risk
  • MD5: 5128c34a7de1371b6ee036d9856e6340
  • Run description: Optimized based on the official Indri 2014 baseline run.

uogTrBwf

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: uogTrBwf
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/12/2014
  • Type: automatic
  • Task: risk
  • MD5: de1ff363b82f06145f0479e5992dc363
  • Run description: Risk sensitive learning to rank, optimised for our new risk-sensitive measure based on hypothesis testing, relative to the Indri standard baseline.

uogTrDuax

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: uogTrDuax
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 5f2f84451747d3b31aef95bea946e878
  • Run description: Divergence from Randomess, Fat & QI generated features (64 total), state-of-the-art learning to rank, diversification.

uogTrDwl

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: uogTrDwl
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: a63aba1edbeed7afa08417a5de4e39ea
  • Run description: Divergence from Randomness, Fat & QI generated features (64 total), state-of-the-art learning-to-rank

uogTrDwsts

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: uogTrDwsts
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: risk
  • MD5: 15b1643649a0c757b86da696f4464c3b
  • Run description: Risk sensitive learning to rank, optimised for our new risk-sensitive measure based on hypothesis testing.

uogTrIwa

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: uogTrIwa
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 66a1da9e3154ef0872de8c8eccee872d
  • Run description: Divergence from Independence, Fat & QI generated features (64 total), robust learning-to-rank.

uogTrq1

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: uogTrq1
  • Participant: uogTr
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: risk
  • MD5: 2bccb8f9989e0683ab8d8e400580bfeb
  • Run description: Aims to select the best retrieval approach based on query features such as QPP.

utbase

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: utbase
  • Participant: ut
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 01f716b4924af6ca288346ad2df8cd3d
  • Run description: Running using only anchor text. Rank by linear interpolation language model + length prior.

utexact

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: utexact
  • Participant: ut
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 17521cad5f7164125d8d05c0c48971d1
  • Run description: Running using only anchor text. Rank by nr of times the exact query string matches anchor texts. Additional documents from linear interpolation language model + length prior.

webisWt14axAll

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: webisWt14axAll
  • Participant: BUW
  • Track: Web
  • Year: 2014
  • Submission: 8/13/2014
  • Type: automatic
  • Task: adhoc
  • MD5: a6f203eec6df8c507a96204f14a1d903
  • Run description: Our main idea is to use published and own created axioms for a ranking model. For this run all implemented axioms are used. First an initial ranking with a indri search engine was created and we just consider the top 50 documents. Then every axiom was applied and a summarized reranking was derived. Used axioms: 'TFC1', 'TFC2', 'TFC3', 'TDC', 'LNC1', 'LNC2', 'TF-LNC', 'LB1', 'STMC1', 'OWN1', 'OWN2', 'OWN3', 'OWN4' Axioms are based on term frequency, term discrimination values, length conditions, semantic similarity conditions and proximity

webisWt14axMax

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: webisWt14axMax
  • Participant: BUW
  • Track: Web
  • Year: 2014
  • Submission: 8/13/2014
  • Type: automatic
  • Task: adhoc
  • MD5: a1cfcf444ed4b892ce8a89cce6af6c9e
  • Run description: Our main idea is to use published and own created axioms for a ranking model. For this run selected axioms are used. First an initial ranking with a indri search engine was created and we just consider the top 50 documents. Then every axiom was applied and a summarized reranking was derived. Used axioms: 'TFC3', 'LNC1', 'LNC2', 'LB1', 'OWN2', 'OWN3', 'OWN4' Axiom selection is based on measuring alpha-ndcg@20 and building difference to baseline values. Then the best top 1600 values, sorted descending, from all 2^14 combinations were choosen and combined to this selection. Axioms are based on term frequency, term discrimination values, length conditions, semantic similarity conditions and proximity

webisWt14axSyn

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: webisWt14axSyn
  • Participant: BUW
  • Track: Web
  • Year: 2014
  • Submission: 8/13/2014
  • Type: automatic
  • Task: adhoc
  • MD5: a1cfcf444ed4b892ce8a89cce6af6c9e
  • Run description: For this run selected axioms are used. Our main idea is to use published and own created axioms for a ranking model. First an initial ranking with a indri search engine was created and we just consider the top 50 documents. Then every axiom was applied and a summarized reranking was derived. Used axioms: 'TFC1', 'TDC', 'LNC2', 'LB1', 'OWN2', 'OWN3' Axiom selection is based on measuring alpha-ndcg@20 and building difference to baseline values. Then counting how frequent differences are below 0. Basend on this the best top 1600 values, sorted ascending, from all 2^14 combinations were choosen and combined to this selection. Axioms are based on term frequency, term discrimination values, length conditions, semantic similarity conditions and proximity

wistud.runA

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: wistud.runA
  • Participant: wistud
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: a011b929e54c71ef9034fb6d5e6418c5
  • Run description: basic retrieval system (language modeling) with spam filtering

wistud.runB

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: wistud.runB
  • Participant: wistud
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: ad54680c3b822c9c2e8a58368a61701f
  • Run description: relevance language model implemented, spam filtering

wistud.runC

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: wistud.runC
  • Participant: wistud
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: ad54680c3b822c9c2e8a58368a61701f
  • Run description: basic deduplication

wistud.runD

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: wistud.runD
  • Participant: wistud
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: risk
  • MD5: f63029594f1395899b8bfdf013fe157a
  • Run description: basic deduplication

wistud.runE

Results | Participants | Input | Summary (std-gd) | Summary (std-nd) | Summary (risk-rm-a0-gd) | Summary (risk-rm-a0-nd) | Summary (risk-rm-a5-gd) | Summary (risk-rm-a5-nd) | Summary (risk-terrier-a0-gd) | Summary (risk-terrier-a0-nd) | Summary (risk-terrier-a5-gd) | Summary (risk-terrier-a5-nd) | Appendix

  • Run ID: wistud.runE
  • Participant: wistud
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: risk
  • MD5: 55839f71922f21eef14fa0781316060c
  • Run description: relevance model applied after cleaning of the initial run

Zerg

Results | Participants | Proceedings | Input | Summary (std-gd) | Summary (std-nd) | Appendix

  • Run ID: Zerg
  • Participant: Group.Xu
  • Track: Web
  • Year: 2014
  • Submission: 8/11/2014
  • Type: automatic
  • Task: adhoc
  • MD5: 61a5ce474f422ce763cd1d4c0bed7d5c
  • Run description: Simple combination of existing LM based systems with baselines. NOT LETOR.