Skip to content

Runs - Decision 2019

ICTNETv1BM25

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: ICTNETv1BM25
  • Participant: ICTNET
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 7c29ba0024c1011bcf3eefe60ad2410f
  • Run description: We extract the content of the document in ClueWeb2_13, convert it into an xml file type, and then use BM25 for retrieval.

ICTNETv2BM25

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: ICTNETv2BM25
  • Participant: ICTNET
  • Track: Decision
  • Year: 2019
  • Submission: 8/28/2019
  • Type: auto
  • MD5: ea99347ee90002688da206d6915d8880
  • Run description: Try method fusion.

IELAB01_ori_q

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB01_ori_q
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: a363b0f21b63e6c6fd7f5f46bf8f2a9b
  • Run description: Use the query field to retrieve results using BM25F on Elasticsearch

IELAB02_ori_d

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB02_ori_d
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: b95687615661a7202057d4102dd6113c
  • Run description: Use the description field to retrieve results using BM25F on Elasticsearch

IELAB03_umls_d

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB03_umls_d
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 9dc6bcc766f09cc13c79075b32c6fdf5
  • Run description: Use the UMLS nGram found in the description field to retrieve results using BM25F on Elasticsearch

IELAB04_umls_n

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB04_umls_n
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: c94a309293fde1cea576c1aeb8f4142a
  • Run description: Use the UMLS nGram found in the narrative field to retrieve results using BM25F on Elasticsearch

IELAB05_xChv_q

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB05_xChv_q
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 986a3498246952792c8cc41c76fc3d7a
  • Run description: Use the EQFE model to expand the query field using terms from CHV and then retrieve results using BM25F on Elasticsearch

IELAB06_xChv_d

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB06_xChv_d
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 072793a6ef389e4663bde13558329210
  • Run description: Use the EQFE model to expand the description field using terms from CHV and then retrieve results using BM25F on Elasticsearch

IELAB07_xWiki_q

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB07_xWiki_q
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 9343953c6fb2fca05d606cc56c4e519f
  • Run description: Use the EQFE model to expand the query field using terms from Wikipedia and then retrieve results using BM25F on Elasticsearch

IELAB08_xWiki_d

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB08_xWiki_d
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: d678d13076635c8b1033ac1fce6247aa
  • Run description: Use the EQFE model to expand the descriotion field using terms from Wikipedia and then retrieve results using BM25F on Elasticsearch

IELAB09_xCW_q

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB09_xCW_q
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: a52fcc08267d260d422257c6b34fe83b
  • Run description: Use the EQFE model to expand the query field using terms from CHV and Wikipedia, and then retrieve results using BM25F on Elasticsearch

IELAB10_xCW_d

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: IELAB10_xCW_d
  • Participant: UQ
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 4a7d285902cbfd93c7557dace2353318
  • Run description: Use the EQFE model to expand the description field using terms from CHV and Wikipedia, and then retrieve results using BM25F on Elasticsearch

UWaterMDS_BM25

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWaterMDS_BM25
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 3e396033756d5a6658a163dc11675771
  • Run description: Baseline Ansereni BM25.

UWatMDS_BM25_Z

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BM25_Z
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 8d2261371d8ae47d67ad24081be733b4
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDS_BM25_ZS

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BM25_ZS
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 064671d26fea1b1adac0296687503b50
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDS_BMF_C90

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BMF_C90
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 55a419f9e395e00238ac637e6c825daf
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDS_BMF_C95

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BMF_C95
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 4d96e251f9078c2b3c53995046af61ac
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDS_BMF_S30

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BMF_S30
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 68afb6d300b7c5540f730d306fc3188b
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDS_BMZBS10

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDS_BMZBS10
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: auto
  • MD5: 0a30bfb05cfcfe70a011f2bc067fa46e
  • Run description: Reranking BM25 results using a credibility classifier or/and spam classifier.

UWatMDSBM25_HC1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDSBM25_HC1
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: manual
  • MD5: 93abe619d4b5f487672e03fed86d2b0a
  • Run description: Used HiCAL to find relevant documents, then return remaining unjudged BM25 documents.

UWatMDSBM25_HC2

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDSBM25_HC2
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: manual
  • MD5: 28463495a2022d2c93d82ce0bfe12e64
  • Run description: Used HiCAL to find relevant documents, then return remaining unjudged BM25 documents, with documents ranking pushed if classifier also found the same document.

UWatMDSBM25_HC3

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: UWatMDSBM25_HC3
  • Participant: UWaterlooMDS
  • Track: Decision
  • Year: 2019
  • Submission: 8/26/2019
  • Type: manual
  • MD5: a6ceed023bec5875ae437f17b70f0f53
  • Run description: Used HiCAL to find relevant documents, then return remaining unjudged BM25 documents, with a document ranking pushed if the classifier also found it. Here, the classifier was trained on a cleaner (only html body, no tag) collection.

webisMAll1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMAll1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/28/2019
  • Type: auto
  • MD5: 6ad2154658d3391f765c75bff5a8016c
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. Re-ranking is only performed when all the five axioms agree together. The argument mining strategy is the same as that used in runs with identifiers with webisMSame and identifier webisMMajority1.

webisMMajority1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMMajority1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/28/2019
  • Type: auto
  • MD5: 5cdf8a7c8c8721a967e0cf0feb48be12
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. Re-ranking is only performed when majority of the axioms (3 out of 5 axioms) agree together. The argument mining strategy is the same as that used in runs with identifiers with webisMSame.

webisMSame1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMSame1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 61c6f8fec8f2dd98fe3afbc2c2762e56
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. All of the axioms have been assigned the same weight.

webisMSame2

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMSame2
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 1c76b10e116a6d34cf6271608a9d9d3b
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. All of the axioms have been assigned the same weight. Finally, we re-rank again with different chunks.

webisMSame3

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMSame3
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: f88f9a78b95138233d7ca645f1e5b168
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. All of the axioms have been assigned the same weight. Finally, we re-rank again with different chunks.

webisMSame4

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMSame4
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 808bc83cd7bd59a4acb66cf0f889f12a
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. All of the axioms have been assigned the same weight. Finally, we re-rank again with different chunks.

webisMSame5

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisMSame5
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 3bf34564cb11e53686534ca550e33271
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. All of the axioms have been assigned the same weight. Finally, we re-rank again with different chunks.

webisTAll1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisTAll1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/28/2019
  • Type: auto
  • MD5: 9e93cbe4adee82df356039ed529006a6
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. Re-ranking is only performed when all the five axioms agree together. The argument mining strategy is the same as that used in runs with identifiers webisTSame1 and webisTMajority1.

webisTMajority1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisTMajority1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/28/2019
  • Type: auto
  • MD5: 29d317a0302e2e4b2638b29fea82b8da
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. Re-ranking is only performed when majority of the axioms (3 out of 5 axioms) agree together. The argument mining strategy is the same as that used in runs with identifier webisTSame1.

webisTSame1

Results | Participants | Proceedings | Input | Summary (trec_eval) | Summary (extended) | Appendix

  • Run ID: webisTSame1
  • Participant: Webis
  • Track: Decision
  • Year: 2019
  • Submission: 8/27/2019
  • Type: auto
  • MD5: 04ce5583b1d3a0cdb1b6e744b49cd5ad
  • Run description: First we apply baseline retrieval function BM25 using Elasticsearch with pre-ranking based on the domain category. For the documents obtained from that, the top-50 documents are re-ranked based on 5 axioms. We use a different argument mining approach than one used in the previous runs.