Skip to content

Runs - Session 2014

ecnusession1.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession1.RL1
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL1
  • MD5: dcf8793abaf43e3b928dce1064148e45
  • Run description: combine terrier BM25 + baseline ecnusession1.RL1

ecnusession1.RL2

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession1.RL2
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL2
  • MD5: 7c1302a014e9341fb798475d1dd10c03
  • Run description: tfidf_click_indri_weight

ecnusession1.RL3

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession1.RL3
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: manual
  • Task: RL3
  • MD5: e07fc8fc1146e677a00fca0994f346e1
  • Run description: manual change + weak rule RL3 based on ecnusession1.RL2

ecnusession2.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession2.RL1
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL1
  • MD5: d741ef3dae6cfbbaea8ce9d8e71d5796
  • Run description: terrier BM25 combine baseline

ecnusession2.RL2

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession2.RL2
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL2
  • MD5: add52a5c0f2a201a9e68ce7280da33c4
  • Run description: LDA + click_indri_weighte_ecnusessio2

ecnusession2.RL3

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession2.RL3
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL3
  • MD5: 7ae431c0ad5d1f82a861af735d5758ed
  • Run description: combine tfidf_click_indri_weighte + Rl3.py ecnusession2.RL3

ecnusession3.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession3.RL1
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL1
  • MD5: 480b62158ff8c2f31f8c2f1b4fe44b73
  • Run description: indri online combine baseline

ecnusession3.RL2

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession3.RL2
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL2
  • MD5: e0b2425f83df79961faeaca6bf4d64a7
  • Run description: svm + features

ecnusession3.RL3

Results | Participants | Input | Summary | Appendix

  • Run ID: ecnusession3.RL3
  • Participant: ecnu
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL3
  • MD5: ad3abfc63c495f01f5b7f02ff6fd5299
  • Run description: combine svm + features + tfidf_indri_weight + RL3.py

ECxCGxPRF.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxCGxPRF.RL1
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: b6baee7c88d085f9a1ce06bfe71105cb
  • Run description: Used 2013 data to generate a document click collocation graph. A random walk over that graph was performed using the top document retrieved for the current query of each session (from the track-provided baseline retrieval) to generate document recommendations. Recommendations that were retrieved in the top 50 ranks of the baseline retrieval results were boosted to the top of the ranking.

ECxCGxPRF.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxCGxPRF.RL2
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL2
  • MD5: 960f72831c65c039bea51faa5fc412e4
  • Run description: Used 2013 data to generate a document click collocation graph. A random walk over that graph was performed using either a) the top document retrieved for the current query of each session (from the track-provided baseline retrieval) or b) clicked documents to generate document recommendations. A) was used for sessions with no clicks. Recommendations that were retrieved in the top 50 ranks of the baseline retrieval results were boosted to the top of the ranking.

ECxCGxPRF.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxCGxPRF.RL3
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL3
  • MD5: e09aed6dc7af4df488eba5087af44db8
  • Run description: Used 2013 and 2014 data to generate a document click collocation graph. A random walk over that graph was performed using either a) the top document retrieved for the current query of each session (from the track-provided baseline retrieval) or b) clicked documents to generate document recommendations. A) was used for sessions with no clicks. Recommendations that were retrieved in the top 50 ranks of the baseline retrieval results were boosted to the top of the ranking.

ECxSRMxOS.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxOS.RL1
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: 36d29e2f2c1cc9c4727c40fd46f27325
  • Run description: This is the baseline run with spam document removed (using a Waterloo spam cutoff of 65).

ECxSRMxOS.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxOS.RL2
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL2
  • MD5: 9a055e50e697dc64e9c5eac7a1206d3b
  • Run description: This is a re-ranking of the baseline run with spam document removed (using a Waterloo spam cutoff of 65). Re-ranking was performed only over queries with clicks. The snippets of clicked documents within a session were used to build a relevance model, and the results of the current query were re-ranked using their snippet text.

ECxSRMxOS.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxOS.RL3
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL3
  • MD5: b971e8ca3abc35881b4c8fc449fd2e62
  • Run description: This is a re-ranking of the baseline run with spam document removed (using a Waterloo spam cutoff of 65). An initial re-ranking was performed only over queries with clicks. The snippets of clicked documents within a session were used to build a relevance model, and the results of the current query were re-ranked using their snippet text. A second re-ranking was performed, boosting documents in the top 50 results that were recommended using a document click collocation graph generated over the 2013 and 2014 data.

ECxSRMxPRF.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxPRF.RL1
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: 0721e9755e1f0028e5bfa191793d3400
  • Run description: This is a re-ranking of the baseline run with spam document removed (using a Waterloo spam cutoff of 65). The snippet of the top document retrieved for the current query of each session was used to create a relevance model, and documents were re-ranked by their snippet text.

ECxSRMxPRF.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxPRF.RL2
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL2
  • MD5: 6cad1b613c77c611a227f33f356d8697
  • Run description: This is a re-ranking of the baseline run with spam document removed (using a Waterloo spam cutoff of 65). For sessions without clicks, the snippet of the top document retrieved for the current query of each session was used to create a relevance model, and documents were re-ranked by their snippet text. For sessions with clicks, the snippet text of clicked documents was used to create the relevance model.

ECxSRMxPRF.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ECxSRMxPRF.RL3
  • Participant: Endicott
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL3
  • MD5: 1936e6c217a9d19e158d138cc386c106
  • Run description: This is a re-ranking of the baseline run with spam document removed (using a Waterloo spam cutoff of 65). For sessions without clicks, the snippet of the top document retrieved for the current query of each session was used to create a relevance model, and documents were re-ranked by their snippet text. For sessions with clicks, the snippet text of clicked documents was used to create the relevance model. A final re-ranking was performed, boosting documents in the top 50 results that were recommended using a document click collocation graph generated over the 2013 and 2014 data.

GUS14Run1.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run1.RL1
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 46847e026bf5f2c847a53fd302d00357
  • Run description: Click docs + CatB Adhoc Search with Lemur

GUS14Run1.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run1.RL2
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: e2517f735e5413311de1ef63930a7a7c
  • Run description: Clicked docs + catB QCM with bad queries weight 0.65 increase clicked documents in a session

GUS14Run1.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run1.RL3
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: 1ff2071f9711ee60396dd3d674d19585
  • Run description: CatB + Clicked documents QCM with 0.65 weighted bad queries session clustred using k-means

GUS14Run2.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run2.RL1
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 7639c08f29c913a2fdae1082d2528a6a
  • Run description: Clicked docs and CatB Adhoc Search using Lemur

GUS14Run2.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run2.RL2
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: e631ec6fed49e2c4c9ec68cca80e4ce1
  • Run description: CatB and clicked docs QCM with bad query weight 0.8 increase clicked documents' score

GUS14Run2.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run2.RL3
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: 7c6f6193d620ea361cde3ad5f85738b9
  • Run description: Click docs and ClueWeb12CatB qcm with bad query weight 0.8 session clustered using k-mean

GUS14Run3.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run3.RL1
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: be22afb6564178da3df4cebdeb52a4ae
  • Run description: clicked docs and CatB adhoc lemur retrieval

GUS14Run3.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run3.RL2
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: fecba96a14c05daf52193cb2f99c0c14
  • Run description: CatB and clicked docs QCM with 0.8 weight for bad queries clicked documents' score increased

GUS14Run3.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: GUS14Run3.RL3
  • Participant: Georgetown
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: d5f024ad2135a7dd072da2a49d962789
  • Run description: clicked docs + CatB QCM + 0.8 weight bad queries used topic number as topic id to organize sessions

ICTNET14SER1.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER1.RL1
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL1
  • MD5: 1d4c7bb40f293acce0a874dde5c958c3
  • Run description: a re-ordered baseline RL1 run

ICTNET14SER1.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER1.RL2
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL2
  • MD5: da0eff30a51b8aa2a6b2fa39dd8fe7ac
  • Run description: ICTNET14SER1.RL2 uses 7 features to train the LTR model which will be used in re-rank the original document order in baseline. The seven features are bm25 score between query and content, cosin similarity between query and title, query expansion score, visual document score, pagerank score, whether the document is clicked and the predicted attention time of the score.

ICTNET14SER1.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER1.RL3
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL3
  • MD5: c14df23a89256676697fae01a4a6b656
  • Run description: ICTNET14SER1.RL3 uses the completely same features of the ICTNET14SER1.RL2, except that ICTNET14SER1.RL3 uses all the information of sessions that have the same topic with the current session.

ICTNET14SER2.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER2.RL1
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL1
  • MD5: 2e1afc7827c1d12d07a25967884e52d1
  • Run description: a re-ordered baseline RL1 run

ICTNET14SER2.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER2.RL2
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL2
  • MD5: fe1728b1fe965bce9c413c30731a82ec
  • Run description: ICTNET14SER2.RL2 uses 4 features to train the LTR model which will be used in re-rank the original document order in baseline. The four features are bm25 score between query and content, cosin similarity between query and title, query expansion score, visual document score.

ICTNET14SER2.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER2.RL3
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL3
  • MD5: 3e9a01f5a2ba0e23bc91a43c5d50327d
  • Run description: ICTNET14SER2.RL3 uses the completely same features of the ICTNET14SER2.RL2, except that ICTNET14SER2.RL3 uses all the information of sessions that have the same topic with the current session.

ICTNET14SER3.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER3.RL1
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL1
  • MD5: 0477409352ad8017963bf80c4bd09199
  • Run description: a re-ordered baseline RL1 run

ICTNET14SER3.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER3.RL2
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL2
  • MD5: 0df6e14d2727dd79da21519a7f0ff295
  • Run description: ICTNET14SER3.RL2 uses 4 features to train the LTR model which will be used in re-rank the original document order in baseline. The four features are cosin similarity between query and title, query expansion score, visual document score and the predicted attention time of the score.

ICTNET14SER3.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: ICTNET14SER3.RL3
  • Participant: ICTNET
  • Track: Session
  • Year: 2014
  • Submission: 9/9/2014
  • Type: automatic
  • Task: RL3
  • MD5: a7e7aea2f3fdfa01d5e76242a8741491
  • Run description: ICTNET14SER3.RL3 uses the completely same features of the ICTNET14SER3.RL2, except that ICTNET14SER3.RL3 uses all the information of sessions that have the same topic with the current session.

RAMA.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: RAMA.RL1
  • Participant: RAMA
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 383b28ee7404057ad134b8f260acf0bd
  • Run description: This RL1 run is based on basic Lucene search with the current query.

RAMA.RL2

Results | Participants | Input | Summary | Appendix

  • Run ID: RAMA.RL2
  • Participant: RAMA
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: 2eba631f9af158a4abb90a5c41b8eab2
  • Run description: This RL2 run is based on a user's interest model built with the data in the same session as the current query. The top-weighted elements in the model is combined with the current query for a refined Lucene search.

RAMA.RL3

Results | Participants | Input | Summary | Appendix

  • Run ID: RAMA.RL3
  • Participant: RAMA
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: 3a6c35394a2b0ecd2aabd10868c7588e
  • Run description: This RL3 run is based on a user's long-term interest model built with the data in all sessions belong to the user. The top-weighted elements in the model is then combined with the current query for a refined Lucene search.

SCIAITeamC.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamC.RL1
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL1
  • MD5: 11ba880e9be8a2d1712d2e9af80b4872
  • Run description: Used the content of the clicked on pages to expand on the query

SCIAITeamC.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamC.RL2
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL2
  • MD5: fc368afc74e8295ea755e44805c0aa04
  • Run description: Used the content of the clicked on pages to expand on the query

SCIAITeamF.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamF.RL1
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL1
  • MD5: 97092ac16b2854e124abf10bcef29109
  • Run description: Used a Framing technique on the "query" and the returned web pages to automatically judge relevance. Named Entity recognition and part of speech tagging was utilized.

SCIAITeamF.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamF.RL2
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL2
  • MD5: 0f0e50cefcaa1e54238856acecd43254
  • Run description: Used a Framing technique on the "query" and the returned web pages to automatically judge relevance. Named Entity recognition and part of speech tagging was utilized.

SCIAITeamL.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamL.RL1
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL1
  • MD5: 068de8cb35ed39b78aa385b6d0591f9a
  • Run description: Used the content of the pages with the longest visited time to expand on the query

SCIAITeamL.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCIAITeamL.RL2
  • Participant: SCIAITeam
  • Track: Session
  • Year: 2014
  • Submission: 8/26/2014
  • Type: automatic
  • Task: RL2
  • MD5: 13904e4b7947d509c4e7236520833406
  • Run description: Used the content of the pages with the longest visited time to expand on the query

uclbaseline.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: uclbaseline.RL1
  • Participant: UCL
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 534cbf7bdecfcc2eaf5edfaa0655187e
  • Run description: Simple baseline retrieval.

udel14Run1.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: udel14Run1.RL1
  • Participant: udel
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 66659cb4dac2e3a0480ec191fb192cc3
  • Run description: Run1 uses a simple data fusion technique, in addition to spam filtering. RL1 does not use information across sessions

udel14Run1.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: udel14Run1.RL3
  • Participant: udel
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: d564f1fd6de3a033427037d160a9342a
  • Run description: Run1 uses a simple data fusion technique, in addition to spam filtering.

udelitu.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: udelitu.RL1
  • Participant: udel
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: d7bbe643e423f1bcc68342716c9acb90
  • Run description: inlink title url

UMASS1.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS1.RL1
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 6e2a8b14037d2f157e2b3f40f0451d6c
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS1.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS1.RL2
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: c8069d88fbb37e40283151afa176cf0e
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS1.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS1.RL3
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: 23549d6513bcfa78746b7e0d12e5d94e
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS2.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS2.RL1
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: c94bd06b41d22e750ae5d56a5bd81eca
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS2.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS2.RL2
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: ecdd95fa95a7cf19e333c83e1b696401
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS2.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS2.RL3
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: c4949880122d6d114a9f1f8d466cae7b
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS3.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS3.RL1
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: 7ecb556a49795e5995990a733244aaaa
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS3.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS3.RL2
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: b18252bbc63877c6951a7d80cc9b04ca
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS3.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS3.RL3
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: b18252bbc63877c6951a7d80cc9b04ca
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS4.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS4.RL1
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL1
  • MD5: eb508fa9148c4a19c1ef315f1bc4b679
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS4.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS4.RL2
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: ab3e243ab4565a11c86b11161b499db5
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

UMASS4.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UMASS4.RL3
  • Participant: UMASS_CIIR
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: a20935716d4093f89850a5b98dafe09a
  • Run description: Learning to rank using: current query, query reformulation, clicked snippet and skipped snippet, SAT clicks and DSAT clicks.

webis2014act.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014act.RL1
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL1
  • MD5: ccea4fb96c34d0622e9707f575931777
  • Run description: Baseline ranking provided by TREC

webis2014act.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014act.RL2
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL2
  • MD5: 72df3f5eb56d8a14b6fb0734d19e49e3
  • Run description: Removed previously viewed documents from baseline ranking.

webis2014act.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014act.RL3
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL3
  • MD5: aea922fbb9e5b2446e648d08794510ce
  • Run description: Reranking according to the spreading activation, based on PMI between phrases of the task description and result's snippets.

webis2014db.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014db.RL1
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: d043286919373bd73aff3b19c89deddc
  • Run description: Baseline for webis2014db

webis2014db.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014db.RL2
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL2
  • MD5: 281e3675efac32b15fe9607f30c95c83
  • Run description: Baseline ranking with previously clicked results removed

webis2014db.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webis2014db.RL3
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/8/2014
  • Type: automatic
  • Task: RL3
  • MD5: 47977eba5df262fcb28cac850409de29
  • Run description: Baseline re-ranked, based on relevance judgements of the previous years.

webisSt14ax.RL1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webisSt14ax.RL1
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: 4debeac5c123401b0a48a88228ebf172
  • Run description: baseline for axiomatic approach

webisSt14ax.RL2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webisSt14ax.RL2
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL2
  • MD5: 7d14bee50af4eb1c6543d7c041a19580
  • Run description: RL2 with concluded axiom weights based on clicked pages and 'syn' weights

webisSt14ax.RL3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: webisSt14ax.RL3
  • Participant: BUW
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL3
  • MD5: 80e6654dd5b27cb55a5f59f59c1287ae
  • Run description: RL3 with concluded axiom weights based on clickresults and 'max' approach

WLZNTJU.RL1

Results | Participants | Input | Summary | Appendix

  • Run ID: WLZNTJU.RL1
  • Participant: wlzn_tju_cn
  • Track: Session
  • Year: 2014
  • Submission: 9/10/2014
  • Type: automatic
  • Task: RL1
  • MD5: 682f7ee77705b7041d5cbb25257d7697
  • Run description: This is a baseline run simply using the LM running on cluewebB.

WLZNTJU.RL2

Results | Participants | Input | Summary | Appendix

  • Run ID: WLZNTJU.RL2
  • Participant: wlzn_tju_cn
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL2
  • MD5: 33add32a710a4bf915c8b1fd486a2903
  • Run description: use only the same session features to train an learning to rank model. we only submit 50 documents for each session for the reason that we have no sufficient time to return 2000 documents.

WLZNTJU.RL3

Results | Participants | Input | Summary | Appendix

  • Run ID: WLZNTJU.RL3
  • Participant: wlzn_tju_cn
  • Track: Session
  • Year: 2014
  • Submission: 9/11/2014
  • Type: automatic
  • Task: RL3
  • MD5: 820eb8e17068b0d66841de76c3f854a8
  • Run description: this result is obtained using learning to rank methods. the features are extracted from current query, previous query, the whole session and the other sessions within the same topic. The click-through data is used as the relevance judgement in labeling. As we run on the cluewebB, so we didn't use the full text of the interaction and only use the title and snippets.