Skip to content

Runs - Chemical 2009

BiTeM09

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 2f21024e959e73ca7f19e102d1fe8760
  • Run description: Document representation One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are : title, abstract, and claims. Indexing We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed without Porter stemming and without query expansion.

BiTeM09comb

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09comb
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 0aec448fae847dd63e8eadd5607d2e0d
  • Run description: This run is a linear combination of maxiStem.res and maxi.res (scoreStem + 0.1*maxi.res).

BiTeM09ipc1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09ipc1
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: d5178ea7d4215f67968f127f680e39e9
  • Run description: Document representation : One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are title, abstract, and claims. Indexing : We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names. Finally, we used our IPC categorizer in order to extract 1 IPC code from query, and to downweight retrieved document that did not share at least one code with the query.

BiTeM09ipc3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09ipc3
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: f4196b6497a13a2facc73b4aa3fc9094
  • Run description: Document representation : One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are title, abstract, and claims. Indexing : We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names. Finally, we used our IPC categorizer in order to extract 3 IPC codes from query, and to downweight retrieved document that did not share at least one code with the query.

BiTeM09ipc3b

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09ipc3b
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: f5fca4926f70b87c6b6e92ef6a44b827
  • Run description: Document representation : One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are title, abstract, and claims. Indexing : We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names. Finally, we used our IPC categorizer in order to extract 3 IPC codes from query, and to downweight (factor 0.9 instead of 0.8) retrieved document that did not share at least one code with the query.

BiTeM09ipc5

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09ipc5
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: aee0c6f919517b7af46bab2ca6544239
  • Run description: Document representation : One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are title, abstract, and claims. Indexing : We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names. Finally, we used our IPC categorizer in order to extract 5 IPC codes from query, and to downweight retrieved document that did not share at least one code with the query.

BiTeM09PAbl

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAbl
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: c217e40701fba7f9dbbac61366e98ece
  • Run description: baseline run

BiTeM09PAcba

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAcba
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: fa4fb57c986a61a24879fd1db1d8874f
  • Run description: mix between cit and cl run with alpha 0.1

BiTeM09PAcbb

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAcbb
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: 52852a4d1e32954fb990473665a3a05f
  • Run description: same as cba with alpha = 1

BiTeM09PAcit

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAcit
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: 6aef402fc4af6b6b192390458ff5de21
  • Run description: post processing with direct citations

BiTeM09PAcl

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAcl
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: a2b142bbf5059a8fe667490edbd4531d
  • Run description: post processing reranking based on claims

BiTeM09PAqe

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: BiTeM09PAqe
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: 91285ae6e87b2c25e1525d39c2441469
  • Run description: post processing strategy with direct citations chemical query expansion strategy with Oscar3 and MeSH

BiTeM09po

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09po
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 8c990c6b9983e3053b3191d284eed3fb
  • Run description: Document representation One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are : title, abstract, and claims. Indexing We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming and without query expansion.

BiTeM09qe

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09qe
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: fef7b806d80c3d5def403558a7cf7654
  • Run description: Document representation One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are : title, abstract, and claims. Indexing We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed without Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names.

BiTeM09qepo

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: BiTeM09qepo
  • Participant: BITEM
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 0266acf0c3b517f9c55f4575ef023270
  • Run description: Document representation One file is created for each patent family. We chose to aggregate all informations we found in patent files, that are : title, abstract, and claims. Indexing We used a vector space search engine in order to index the collection. We used a combination of weighting models. This runs was computed with Porter stemming. We performed query expansion relying on Pubchem and Oscar3 in order to expand chemical names.

CMU09Chmtcd

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: CMU09Chmtcd
  • Participant: CMU_LIRA
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: cbe324e38913dca0110bce51112dd2ed
  • Run description: title claim weighted combined, date filtered (result publication date should be before the latest priority date of the query).

CMU09Chmtcdd

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: CMU09Chmtcdd
  • Participant: CMU_LIRA
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 4111bcf2de32f0c60c1141bdfd305ddd
  • Run description: title + claim (words in claim are weighted by occurrence counts in description and abstract), date filtered (result publication date should be before the latest priority date of the query).

DUT09TSRun1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUT09TSRun1
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 5c8aaeb1e61efc857e8fb1aac42484ec
  • Run description: The results of original query

DUT09TSRun2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUT09TSRun2
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: c23d4c28801c257068a0900b605b8350
  • Run description: Using the narrative information.

DUT09TSRun3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUT09TSRun3
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 16ea4e80f9e3ef42798f793676bb53b5
  • Run description: Using the narrative information and feedback.

DUT09TSRun4

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUT09TSRun4
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 41427a7dacba12edde3a9fcd20bc22e2
  • Run description: separately ranking each fields based on In_expB2 weighting model, and then combine the scores linearly.

DUT09TSRun6

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUT09TSRun6
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 61061999f654bb8c390e107aff02935e
  • Run description: SIMPLE BM25F like wighting model, but the weighting function is In_expB2

DUTIR09BM25F

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUTIR09BM25F
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: f3c7f3090bbaa45b32f765d557eb47a2
  • Run description: using bm25f structure-based weighting model.

DUTIRRun1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUTIRRun1
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: dd2b43de912bd1785a58e406bbd3bd32
  • Run description: we use the title,abstract,description,claims field of the patents,and select 60 terms to express each patent.

DUTIRRun2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUTIRRun2
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 275a40664f3ca7fb2789b0fe68190730
  • Run description: we use the title,abstract,description,claims field of the patents,and select 60 terms weighted to express each patent.

DUTIRRun3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: DUTIRRun3
  • Participant: DUTIR
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 7edcd5e445fc1c7c9d3af05e6c2c6b09
  • Run description: We use the title,abstract,description,claims field of the patents,and select 60 terms weighted to express each patent. And rerank the results using some additional terms.

msoe09TSx1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx1
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 5a1c473398bb0d383478770dcfe24e03
  • Run description: // concept document, concept paragraph, document sc, document all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx1" "aggdocresultssc" "(1conceptdoccountnorm + 0.1conceptdocidfsumnorm + 0.01conceptmaxparcountnorm + 0.001conceptmaxparidfsumnorm + 0.0001docscnorm + 0.00001docnormscall )" "msoe09TSx1" "msoe09TSx1"

msoe09TSx1ta

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx1ta
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 04834691b7ae964b59a1561b6601cdcc
  • Run description: "(1conceptdoccountnorm + 1conceptdocidfsumnorm + 0.1conceptmaxparcountnorm + 0.1conceptmaxparidfsumnorm + 0.01docscnorm + 0.01maxParScNorm)"

msoe09TSx2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx2
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: c2a3b363f507502cd7df80a56ce11729
  • Run description: // concept document, concept paragraph, document sc, document all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx1" // concept paragraph, concept document, document sc, paragraph sc, document all sc, paragraph all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx2" "aggdocresultssc" "(0.01conceptdoccountnorm + 0.001conceptdocidfsumnorm + 1conceptmaxparcountnorm + 0.1conceptmaxparidfsumnorm + 0.0001docscnorm + 0.00001docnormscall )" "msoe09TSx2" "msoe09TSx2"

msoe09TSx2ta

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx2ta
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: cc0b8d3c50c99df5d72dc50220dc22a9
  • Run description: "(0.1conceptdoccountnorm + 0.1conceptdocidfsumnorm + 1conceptmaxparcountnorm + 1conceptmaxparidfsumnorm + 0.01docscnorm + 0.01maxParScNorm )"

msoe09TSx3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx3
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: e278f2e7542dff649aa240fc3490448e
  • Run description: // concept document, concept paragraph, document sc, document all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx1" // concept paragraph, concept document, document sc, paragraph sc, document all sc, paragraph all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" // equal waiting all models sans doc all "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx3" "aggdocresultssc" "(1conceptdoccountnorm + 1conceptdocidfsumnorm + 1conceptmaxparcountnorm + 1conceptmaxparidfsumnorm + 1docscnorm + 0docnormscall )" "msoe09TSx3" "msoe09TSx3"

msoe09TSx3ta

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx3ta
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: e063b585271d07e1883f10608f8bf0c4
  • Run description: "(1conceptdoccountnorm + 1conceptdocidfsumnorm + 1conceptmaxparcountnorm + 1conceptmaxparidfsumnorm + 1docscnorm + 1maxParScNorm )"

msoe09TSx4

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx4
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 12bdbee12aa46faac2bb8020ae0a7e78
  • Run description: // concept document, concept paragraph, document sc, document all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx1" // concept paragraph, concept document, document sc, paragraph sc, document all sc, paragraph all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" // equal waiting all models sans doc all "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" // no concepts, document sc, document all sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx4" "aggdocresultssc" "(0conceptdoccountnorm + 0conceptdocidfsumnorm + 0.0conceptmaxparcountnorm + 0.0conceptmaxparidfsumnorm + 1.0docscnorm + 0.1docnormscall )" "msoe09TSx4" "msoe09TSx4"

msoe09TSx4ta

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx4ta
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 468d9484072c823f8551f8981e167b2b
  • Run description: "(0.01conceptdoccountnorm + 0.01conceptdocidfsumnorm + 0.001conceptmaxparcountnorm + 0.001conceptmaxparidfsumnorm + 1.0docscnorm + 0.1maxParScNorm )"

msoe09TSx5

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx5
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: c03de2bf0b495f642f4315eb2e218e38
  • Run description: // concept document, concept paragraph, document sc, document all sc // concept document n, concept paragraph n, document sc "jdbc:oracle:thin:scott/tiger@localhost:1521:orcl" "trecchemresultsx5" "aggdocresultssc" "(1conceptdoccountnorm + 0.0conceptdocidfsumnorm + 0.1conceptmaxparcountnorm + 0.0conceptmaxparidfsumnorm + 0.01docscnorm + 0.0docnormscall )" "msoe09TSx5" "msoe09TSx5"

msoe09TSx5ta

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: msoe09TSx5ta
  • Participant: msoe
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: eb979202361eb00295aa41c9912e3890
  • Run description: "(.001conceptdoccountnorm + 0.001conceptdocidfsumnorm + 0.01conceptmaxparcountnorm + 0.01conceptmaxparidfsumnorm + 0.1docscnorm + 1.0maxParScNorm )"

purduePA09r1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: purduePA09r1
  • Participant: Purdue_Si
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 6f947ac514af7a6b3474eef0e3093369
  • Run description: This is the mandatory run that's expected from all participants: it uses all title words and a selected set of terms from the claims field of query patents (top 50 terms with respect to TF-IDF score are chosen). We have filtered out the documents from the ranked lists, if their minimum priority date is after the maximum priority date of the query file. If priority dates don't exist, we use publication dates as the corresponding dates. We have also exploited the IPC code similarity between query patent and its retrieved patents to rerank the retrieved documents.

purduePA09r2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: purduePA09r2
  • Participant: Purdue_Si
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 35cf985046f92ffa13d91a87d607e2d1
  • Run description: To construct the queries from patent files, we used all title words, and selected sets of terms from the claims, abstract and description fields of query patents (top 30 terms with respect to TF-IDF scores are chosen for each). From those terms, chemical entities are extracted and synonyms for the identified chemical entities are also included from PubChem. Then structured queries are formed to do retrieval over different fields of documents with different weights. Then some of the retrieved documents are filtered out from the ranked list because of date constraints (if their minimum priority date is after the maximum priority date of the query file). If priority dates don't exist, we use publication date as the corresponding dates. Last but not the least, IPC similarities between query patent and its retrieved patents are exploited to rerank the retrieved documents.

purdueTS09r1

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: purdueTS09r1
  • Participant: Purdue_Si
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: cc6cb7668992895a6497c1035e4a9ff3
  • Run description: we used a weighted combination of query title and narrative to do retrieval on different fields of patents and documents (such as title, abstract, claims, description, whole doc etc.) in a weighted way. we used synonyms for chemicals that have been identified, used simple entity recognition to extract information that is later used to increment or decrement weights of some terms, filtering documents from the ranked list etc.

SCAI09PAf1a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf1a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 263f39700b25c82cc49d7677c8f23c85
  • Run description: DOC Vs DOC, Token based search, result filtered

SCAI09PAf1b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf1b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 00206de501b60805e7bb24bfa245455b
  • Run description: Weighted Zone search; token based; result filtered

SCAI09PAf1c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf1c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 01276f721858fb1539929150ba3d2c0e
  • Run description: DESC vs DESC; token based; result filtered

SCAI09PAf1d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf1d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 0e6c9bf547a6475f1c01b51ab7ad39e8
  • Run description: Claim vs Claim; token based; result filtered

SCAI09PAf1e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf1e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 9afadc83cbe5e29fb62daf5e6f31a42e
  • Run description: Doc vs Doc; ipc defined; token based; result filtered

SCAI09PAf2a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf2a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: abc0ad73806cdc0ce9057c0c60a82a25
  • Run description: DOC Vs DOC; NP based search; Results Filtered

SCAI09PAf2b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf2b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: fbe5ea0bc09c654f603fcddc3950bd52
  • Run description: Weighted Zone search; NP Based; Result Filtered

SCAI09PAf2c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf2c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 8467d0e4568acf953525c64e092f4adc
  • Run description: DESC Vs DESC, NP based search, Result Filtered

SCAI09PAf2d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf2d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 892ee2b4b7f43a3329f92d41047d7026
  • Run description: Claim vs claim; NP based; Result Filtered

SCAI09PAf2e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf2e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 71f72f8644e20723aaef2e5a147dcbe4
  • Run description: DOC Vs DOC, NP based search, IPC class defined, Result Filtered

SCAI09PAf3a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf3a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 4d01d2e49b3f81a55c5a2df7269f0646
  • Run description: doc vs doc; concept based; result filtered

SCAI09PAf3b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf3b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: a2feff29c21f26140216a740eaaba42f
  • Run description: weighted zone search; concept based; result filtered

SCAI09PAf3c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf3c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 7143157a32e5e70c5a2a11044bd39ee0
  • Run description: Desc vs Desc; concept based; result filtered

SCAI09PAf3d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf3d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 5e828e1b193f6c4bc4cf7d704f357f9f
  • Run description: claim vs claim; concept based; result filtered

SCAI09PAf3e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf3e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 2346fa6388fd6ffaab703ec5d4dfc796
  • Run description: doc vs doc; ipc defined; concept based; result filtered

SCAI09PAf4a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf4a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 65759dcc73670b643f1d154613399238
  • Run description: Doc Vs Doc; concept based (priority for chemicals); result filtered

SCAI09PAf4b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf4b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 409912b7f39a978cfa2f75cfc355ba98
  • Run description: Doc Vs Doc; Concept based (priority to NPs); Result Filtered

SCAI09PAf4c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAf4c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: baef5eb35cb53b5bec6b7355f84a3051
  • Run description: Doc vs Doc; concept based (priority to related hypernyms and hyponyms); result filtered

SCAI09PAt1a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt1a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 83dcc36ee9a936ec18b4aaf0e9b361f9
  • Run description: DOC Vs DOC, Token based search, Result Unfiltered

SCAI09PAt1b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt1b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: ebe3e3c504835e480a19c1ea9a4bfa27
  • Run description: Weighted zone search, token based, result unfiltered

SCAI09PAt1c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt1c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 9998207a6c4e455da8d264f3961d9b66
  • Run description: Desc Vs Desc, token based, result unfiltered

SCAI09PAt1d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt1d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: f294bdff6297939ac5723df5b35bcdff
  • Run description: claim vs claim; token based; result unfiltered

SCAI09PAt1e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt1e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 9b642199cbab6af5c4f25f9ff4d83a89
  • Run description: Doc Vs Doc, IPC defined, token based; result unfiltered

SCAI09PAt2a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt2a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: f2548aeedafa0eb05f3a539e86409f6d
  • Run description: DOC Vs DOC; NP based search; Unfiltered Result

SCAI09PAt2b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt2b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 98fe10bf7fd9372e37df6ac2cb10b612
  • Run description: Weighted Zone Search; NP Based; Result Unfiltered

SCAI09PAt2c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt2c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: fce9798e2895dcf29a20f099b894eec3
  • Run description: DESC Vs DESC, NP based , Result unfiltered

SCAI09PAt2d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt2d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 7de7be045ab918085802e641761ddc26
  • Run description: Claim vs claim; NP based; Result unfiltered

SCAI09PAt2e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt2e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/5/2009
  • Type: automatic
  • Task: priorart
  • MD5: 290d86edaa506affcb010f948724e5a2
  • Run description: DOC vs DOC; NP based; IPC class defined; Result unfiltered

SCAI09PAt3a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt3a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 225b9929ab0f8ac15bf9eb51ef75f922
  • Run description: DOC vs DOC; concept based; result unfiltered

SCAI09PAt3b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt3b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 5e282f17e0ad6ebe778b84b656af506a
  • Run description: weighted zone search; concept based; result unfiltered

SCAI09PAt3c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt3c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: adb7f6e1fbc4525de346054466b9bf4c
  • Run description: Desc Vs Desc; concept based; result unfiltered

SCAI09PAt3d

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt3d
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: ca8d30195c29a9a6732a3707c211ad2b
  • Run description: claim vs claim; concept based; result unfiltered

SCAI09PAt3e

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt3e
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 207be92b5971dbd329561f392eb911b0
  • Run description: doc vs doc; ipc defined; concept based; result unfiltered

SCAI09PAt4a

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt4a
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: e25857bb8b4fd0063d1a663306de98a7
  • Run description: Doc Vs Doc; Concept Based (priority to chemicals); Result unfiltered

SCAI09PAt4b

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt4b
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: 8c3a23c5c3448b017fc990a89a0af0ae
  • Run description: Doc Vs Doc; Concept Based (priority to NPs); Result unfiltered

SCAI09PAt4c

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: SCAI09PAt4c
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 9/6/2009
  • Type: automatic
  • Task: priorart
  • MD5: b5c165707d0d27783ddd2bae57de327f
  • Run description: Doc vs Doc; concept based (priority to related hyponyms and hypernyms); Result unfiltered

SCAI09TSMAN

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCAI09TSMAN
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: manual
  • Task: techsurv
  • MD5: 9f39f9dd0ffb767e138cded643d5b96a
  • Run description: Manually formulated queries for searching the document index

SCAI09TSNP

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCAI09TSNP
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 7ed4eba648494475dbed0464c0fa58ce
  • Run description: A combination of noun phrase chunking, stop word removal and data cleansing for querying against document index.

SCAI09TSPM

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: SCAI09TSPM
  • Participant: NERCHEM116
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 178e14b89570af8f3c94a49d1781bdc2
  • Run description: Use of a preprocessed dictionary and in-house Named Entity Recognition system for document retrieval

UIowaS09PA1

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: UIowaS09PA1
  • Participant: IowaS
  • Track: Chemical
  • Year: 2009
  • Submission: 9/4/2009
  • Type: automatic
  • Task: priorart
  • MD5: 93a8e5d3da6513d6d13755c9357f359e
  • Run description: We produced 2 separate indexes using Lucene: one using the claims and title information only; the other using other information in the patent including the description, abstract, and classification. We used a specific technique to "un-nest" the claims so they would each be independent documents for evaluation. After limiting the available matching patents using the priority date as a threshold, we merged these two indexes using a weighted ratio and obtained a ranked list of resultant patent documents. These patents were then unduplicated, taking the lowest ranked value of a candidate patent.

UIowaS09PA2

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UIowaS09PA2
  • Participant: IowaS
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 4371efb08001cc4c82dd4d1f40a8c1ab
  • Run description: This run is highly experimental, and used the primary classification and title information to determine an appropriate set of patents to match with, with a heavy weighting on the primary classification. This run allowed the primary patent classification for the target patent to match against the primary classifications for the most appropriate candidate patents. Because priority dates were not used, only an approximation of the earlier patent numbers could be made.

UIowaS09PA3

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: UIowaS09PA3
  • Participant: IowaS
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart_sm
  • MD5: 2718980578125492ce8df52353a396b8
  • Run description: This run is similar to the UIowaS09PA2 run and used the primary classification and title information to determine an appropriate set of patents to match with, with a heavy weighting on the primary classification. This run allowed the primary patent classification for the target patent to match against the primary classifications for the most appropriate candidate patents. This run uses a better approximation of priority dates than UIowaS09PA2, at the expense of returning fewer patents per query.

york09ca02

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca02
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: fe8ad672ab3f68e5b851840b027393e6
  • Run description: We have used Okapi(model: BM25) for retrieval. For documents: Only title and abstract are utilized. For topics: We used query, narrative and some expanded keywords.

york09ca03

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca03
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: e220af53f5bf6734e6bad4fee2e21f09
  • Run description: First, we use the Lucene IR system to build the index for each field. Second, BM25 with default parameter settings is used for weighting.

york09ca04

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca04
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: aede8abdd2f779cba50c019af6a135e7
  • Run description: First, we use the Lucene IR system to build the index for each field. Second, BM25 with default parameter settings is used for weighting. Only the "query" field is used.

york09ca05

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca05
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: c36e51521dcfada87272b664b949a188
  • Run description: Okapi, we only built index based on the title and abstract of the documents.

york09ca06

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca06
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/1/2009
  • Type: automatic
  • Task: techsurv
  • MD5: e5c2761245cb01519b79dfe36e7f98f7
  • Run description: Okapi, we only built index based on the title and abstract of the documents. We only used "query" field of topics to retrive.

york09ca07

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca07
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: f69845471644e086d04ff8b64810f8da
  • Run description: We have used Lucene(model: BM25) for retrieval. For documents: Whole document. For topics: We used query, narrative and some expanded keywords.

york09ca08

Results | Participants | Proceedings | Input | Summary | Appendix

  • Run ID: york09ca08
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 7/8/2009
  • Type: automatic
  • Task: techsurv
  • MD5: 866944d8b7ab32fc10e0f05e01334c40
  • Run description: We have used Lucene(model: Language Model) for retrieval. For documents: Whole document. For topics: We used query, narrative and some expanded keywords.

york09caPA01

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: york09caPA01
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: 09b3570d56638e352ddac6bf46c8abeb
  • Run description: We extracted 30 keywords from title and claim for each topic, and used BM25 as weighting function.

york09caPA03

Results | Participants | Proceedings | Input | Summary (full) | Summary (small) | Appendix

  • Run ID: york09caPA03
  • Participant: york09
  • Track: Chemical
  • Year: 2009
  • Submission: 9/7/2009
  • Type: automatic
  • Task: priorart
  • MD5: 776b778b3f890d157b22eac158555da7
  • Run description: We extracted 100 keywords from title and claim for each topic, and used BM25 as weighting function.