Runs - Web 2012¶
2012bpacad4¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: 2012bpacad4
- Participant: BudapestAcad
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
8242e80eee46c8c6c8cb6b3ad6306016
- Run description: Search run on a distributed architecture using Hadoop and Voldemort. The spammiest 4% of the documents have been removed.
2012bpacad4h¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: 2012bpacad4h
- Participant: BudapestAcad
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
374e09a12236410370b9d5a608a69363
- Run description: Search run on a distributed architecture using Hadoop and Voldemort. The spammiest 4% of the documents have been removed. Host rank has been incorporated into the score.
2012bpacad8¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: 2012bpacad8
- Participant: BudapestAcad
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
93248d459f73fb71ac381a168adafbf1
- Run description: Search run on a distributed architecture using Hadoop and Voldemort. The spammiest 8% of the documents have been removed.
autoSTA¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: autoSTA
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
4363b57f0d9eeeaf199d4961e42e66a7
- Run description: Automatic Subtopics (Subtopics were from wikipedia using the search API, Bing and Yahoo Related Queries API) + a novel diversification function
autoSTB¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: autoSTB
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
d55aa1fa86e6a7258d0909b0875f27d7
- Run description: Automatic Subtopics (Subtopics were from wikipedia using the search API, Bing and Yahoo Related Queries API) + a novel diversification function on CatB
combMNZCatB¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: combMNZCatB
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
628dbe29d8824ac8a94c42f9ac6b4fd2
- Run description: Query Expansion (Subtopics were from wikipedia using the search API, Bing and Yahoo Related Queries API) + combMNZ
DFalah120A¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DFalah120A
- Participant: uottawa
- Track: Web
- Year: 2012
- Submission: 7/17/2012
- Type: automatic
- Task: adhoc
- MD5:
d1cfb1ccf85e526f7d3362f26f11664a
- Run description: In this year, we used our method in last year. For queries that have few results, we build a dictionary from wikipedia topics that have impact terms and 2009-trec one million query topics. The resulted topical dictionary is used for indexing the dataset.
DFalah120D¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DFalah120D
- Participant: uottawa
- Track: Web
- Year: 2012
- Submission: 7/17/2012
- Type: automatic
- Task: diversity
- MD5:
e858c89a2f052fb63a1eddb57b5697fa
- Run description: In this year, we build a dictionary from wikipedia topics that have impact terms and 2009-trec one million query topics. The resulted topical dictionary is used for indexing the dataset.
DFalah121A¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DFalah121A
- Participant: uottawa
- Track: Web
- Year: 2012
- Submission: 7/17/2012
- Type: automatic
- Task: adhoc
- MD5:
f57d24c8af720847e0a97d71b9906279
- Run description: In this year, we used our method in last year. For queries that have few results, we build a dictionary from wikipedia topics that have impact terms and 2009-trec one million query topics. The resulted topical dictionary is used for indexing the dataset.
DFalah121D¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DFalah121D
- Participant: uottawa
- Track: Web
- Year: 2012
- Submission: 7/17/2012
- Type: automatic
- Task: diversity
- MD5:
bcb582d0eeb4c0802b9432e9625e822b
- Run description: In this year, we used our method in last year. For queries that have few results, we build a dictionary from wikipedia topics that have impact terms and 2009-trec one million query topics. The resulted topical dictionary is used for indexing the dataset.
DMthbCatASP¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DMthbCatASP
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
d84f83ad55587acf1a64baf97884788d
- Run description: Markov Random Field Model on Title, Heading and Body on the full ClueWeb09 Collection
DMthbCatB¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: DMthbCatB
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
9754c19148be46ad2b444901ecfff72f
- Run description: Markov Random Field Model on Title, Heading and Body
ICTNET12ADR1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12ADR1
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 7/19/2012
- Type: automatic
- Task: adhoc
- MD5:
e6b4c62ba44d41cd1043ddf8455aef92
- Run description: In run ICTNET12ADR1, we build two index separately. (1)INDEX_WIKI is built for pages in ClueWeb09 under en.wikipedia.org . (2)The other pages in ClueWeb09 is built in INDEX_NORMAL. For INDEX_WIKI, we use Boolean Retrieval Model in url field. For INDEX_NORMAL, the BM25 with sliding window model was used in title and content field to rank the result. The final submission was a combination of the two index above. For each query, (1)We search across INDEX_WIKI first and choose the top 1 result if any matches. (2)Then search across INDEX_NORMAL and fill the other 9999 or 10000 results.
ICTNET12ADR2¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12ADR2
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 8/5/2012
- Type: automatic
- Task: adhoc
- MD5:
b0ca27f91ff258cd674d7420080b6551
- Run description: In run ICTNET12ADR2, we use new html extractor to validate its performance. The other workflow was nearly the same as ICTNET12ADR1. For each query. (1)We search across INDEX_WIKI first and choose the top 1 result if any matches. (2)Then search across INDEX_NORMAL(Title Or, Content And) and fill the other 9999 or 10000 results.
ICTNET12ADR3¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12ADR3
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
a5007f9039a7d796918630a11d825139
- Run description: We use Learning to Rank based on Run 2. RankBoost was used. Features are LM.DIR / LM.JM / LM.KL on title/content/url/fuzzy url, spam value, min_windows_size.
ICTNET12DVR1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12DVR1
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 7/29/2012
- Type: automatic
- Task: diversity
- MD5:
c6eaa5c35e3d0bb071f8fc90f458c58f
- Run description: use LDA to generate the sub-topics and use the xQuAD model to reranking the results
ICTNET12DVR2¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12DVR2
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 8/2/2012
- Type: automatic
- Task: diversity
- MD5:
5ae7af21f5f7be7605c8ac237ed77584
- Run description: Almost the same as run 1, but stemming the contents before LDA.
ICTNET12DVR3¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: ICTNET12DVR3
- Participant: ICTNET
- Track: Web
- Year: 2012
- Submission: 8/7/2012
- Type: automatic
- Task: diversity
- MD5:
6ec1483c03c72e4db3b035dfd40aeb53
- Run description: All the same with run1 but with using different subtopic number and a better normalize ways.
irra12a¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: irra12a
- Participant: IRRA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
ea4a8a83f70b47ad03a8f1e19585dfd4
- Run description: irra12a uses a weighting model based on Divergence From Independence.
irra12b¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: irra12b
- Participant: IRRA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
8928bf2c1555de180e5304ee0e00d7ba
- Run description: irra12b uses a weighting model based on Divergence From Independence.
irra12c¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: irra12c
- Participant: IRRA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
47a178f6bcad155d3cd26bf295a2d37e
- Run description: irra12a uses a weighting model based on Divergence From Independence.
lcm4res¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: lcm4res
- Participant: LIA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: diversity
- MD5:
bab3b9785e73d6418e836fe1d0b10f92
- Run description: Modeling search concepts using a query-oriented algorithm based on Latent Dirichlet Allocation. Concepts are modeled from various sources of information---the ClueWeb09, Wikipedia, the NYT LDC corpus and the GigaWord LDC corpus.
lcmweb¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: lcmweb
- Participant: LIA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
69e84981aaf41c53db35ea5a2f10af11
- Run description: Modeling search concepts using a query-oriented algorithm based on Latent Dirichlet Allocation.
lcmweb10p¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: lcmweb10p
- Participant: LIA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
4e0ed3651bcaa4bf42c04ccbe33f5422
- Run description: Modeling search concepts using a query-oriented algorithm based on Latent Dirichlet Allocation. Concepts are modeled from the 10 best pages retrieved by the query.
lcmwebnoW¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: lcmwebnoW
- Participant: LIA
- Track: Web
- Year: 2012
- Submission: 8/8/2012
- Type: automatic
- Task: adhoc
- MD5:
d8b30e4863ca353464704503fe657e62
- Run description: Modeling search concepts using a query-oriented algorithm based on Latent Dirichlet Allocation. Same run as lcmweb without weighting concepts.
manualSTA¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: manualSTA
- Participant: udel
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: manual
- Task: diversity
- MD5:
e78fdd97b62f632fe880735a86216361
- Run description: Manual Subtopics (Subtopics were created by grad students manually) + a novel diversification function
QUTparaBline¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: QUTparaBline
- Participant: QUT_Para
- Track: Web
- Year: 2012
- Submission: 8/7/2012
- Type: automatic
- Task: adhoc
- MD5:
80bc592a037d22bc2550fa1b070fa5af
- Run description: This is our baseline run that uses google to determine the first number of documents in the returned result set
QUTparaTQEg1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: QUTparaTQEg1
- Participant: QUT_Para
- Track: Web
- Year: 2012
- Submission: 8/3/2012
- Type: automatic
- Task: adhoc
- MD5:
d480bdf707299f3944ce67bf85a10f7f
- Run description: This run uses a query expansion technique that is based on a semantic space model primed from a number of clueweb filtered google results.
qutwa¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: qutwa
- Participant: qutir12
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
d7ce54ef17dc89d78119e75367954ce9
- Run description: TopSig run, no term stats TOPIC-OUTPUT-K = 10000 PSEUDO-FEEDBACK-SAMPLE = 8 Run on Intel i7 950 3.07GHz, 24GB RAM.
qutwb¶
Results
| Participants
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: qutwb
- Participant: qutir12
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
c603d4e2b746ea46709dd5798ac5042f
- Run description: TopSig run, no term stats TOPIC-OUTPUT-K = 10000 PSEUDO-FEEDBACK-SAMPLE = 9 PSEUDO-FEEDBACK-RERANK = 120 Run on Intel i7 950 3.07GHz, 24GB RAM.
srchvrs12c00¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: srchvrs12c00
- Participant: srchvrs
- Track: Web
- Year: 2012
- Submission: 8/6/2012
- Type: automatic
- Task: diversity
- MD5:
b27c579e8030ec72d42fac72ea3afe34
- Run description: Waterloo University spamrank, CMU PageRank
srchvrs12c09¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: srchvrs12c09
- Participant: srchvrs
- Track: Web
- Year: 2012
- Submission: 8/6/2012
- Type: automatic
- Task: adhoc
- MD5:
83b59c196c39dc213f9d5d986ca2c050
- Run description: Waterloo University spamrank, CMU PageRank, MIREX anchortext
srchvrs12c10¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: srchvrs12c10
- Participant: srchvrs
- Track: Web
- Year: 2012
- Submission: 8/6/2012
- Type: automatic
- Task: diversity
- MD5:
adadd5e1f59124ea8c0e11616648525b
- Run description: Waterloo University spamrank, CMU PageRank, MIREX anchortext
UDInfoDivC1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: UDInfoDivC1
- Participant: udel_fang
- Track: Web
- Year: 2012
- Submission: 8/1/2012
- Type: automatic
- Task: diversity
- MD5:
c9409691d4842b37ad942ea98cf0c0b1
- Run description: Combine diversification method using ODP.
UDInfoDivC2¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: UDInfoDivC2
- Participant: udel_fang
- Track: Web
- Year: 2012
- Submission: 8/1/2012
- Type: automatic
- Task: diversity
- MD5:
e1a8bfcd5aaa426e8d2a49a71f6c014d
- Run description: Combine diversification method using DBPedia.
UDInfoDivSt¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: UDInfoDivSt
- Participant: udel_fang
- Track: Web
- Year: 2012
- Submission: 8/1/2012
- Type: automatic
- Task: diversity
- MD5:
e9e7c7d2011637885335d6612319a2f3
- Run description: Adaptive structural diversification methods using DBPedia
uogTrA44s9¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrA44s9
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
051665b3adb3f42ac2ad9c59237a5dc7
- Run description: Learned model including various Divergence from Randomness and other features, and a novel method to increase recall.
uogTrA44xi¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrA44xi
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
cd3b8cb108716580cb5bcc5284f157a3
- Run description: Learned model including various Divergence from Randomness and other features, with improved coverage of related queries using the xQuAD diversification framework.
uogTrA44xl¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrA44xl
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
69eb346628ba295c46f267f20441a78c
- Run description: Diversified ranking produced by a novel machine learning approach to search result diversification, using query- and query-suggestion-dependent features.
uogTrA44xu¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrA44xu
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
35906e25040006b7b92f314fe8decba2
- Run description: Diversified ranking produced by the xQuAD framework, using query suggestions and machine learned models of relevance, coverage, and novelty.
uogTrB44xu¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrB44xu
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
4b199c805454a1f8c96186907b62f0fb
- Run description: Diversified ranking produced by the xQuAD framework, using query suggestions and machine learned models of relevance, coverage, and novelty.
uogTrB45aIs¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: uogTrB45aIs
- Participant: uogTr
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
9d052c3764cf1c0af86fdec29b3f7ff9
- Run description: Learned model including various Divergence from Randomness and other features, including a novel technique to learn the importance of groups of pages.
utw2012c1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012c1
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
144dcae79e8881bcb1d64cce88253dc8
- Run description: Ensemble clustering combination 1
utw2012c2¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012c2
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
051eb819bd8e74a7326e3d60f55358a2
- Run description: ensemble combination 2
utw2012fc1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012fc1
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/10/2012
- Type: automatic
- Task: adhoc
- MD5:
096a7039805852a20f921fba66226de1
- Run description: full ensemble combination
utw2012lda¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012lda
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
9b22f2666421eb4d74a013086ee219c3
- Run description: Diversity baseline based on LDA
utw2012lm09¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012lm09
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: adhoc
- MD5:
229a8eb20a3945aacc0ef811e35d29bc
- Run description: Language model baseline based on anchor text
utw2012sc1¶
Results
| Participants
| Proceedings
| Input
| Summary (trec_eval)
| Summary (ndeval)
| Summary (gdeval)
| Appendix
- Run ID: utw2012sc1
- Participant: utwente
- Track: Web
- Year: 2012
- Submission: 8/9/2012
- Type: automatic
- Task: diversity
- MD5:
9f360e9c67fc585facbf2a78b9d5acd5
- Run description: Simple ensemble combination