Skip to main content

Table 6 IMT micro-averaged participant results

From: The Protein-Protein Interaction tasks of BioCreative III: classification/ranking of articles and linking bio-ontology concepts to full text

Team Run/Srvr Precision Recall F-Score MCC AUC iP/R
Base _Top4 RUN _1 28.60 48.20 35.90 - 0.15924
Base _ Regex RUN _1 32.45 22.34 26.46 - -
T65 RUN_1 8.77 84.82 15.89 0.23552 0.27588
T65 RUN_2 2.45 100.00 4.78 0.06259 0.24484
T65 RUN_3 9.42 81.78 16.89 0.24172 0.27727
T65 RUN_4 33.48 42.32 37.39 0.36166 0.14169
T65 RUN_5 2.44 100.00 4.76 0.06193 0.29016
T69 RUN_1 52.07 55.03 53.51 0.52519 0.34302
T69 RUN_2 54.34 53.51 53.92 0.52958 0.33824
T69 RUN_3 57.36 50.29 53.59 0.52796 0.32539
T69 RUN_4 59.25 48.01 53.04 0.52456 0.31711
T69 RUN_5 61.33 43.64 51.00 0.50896 0.29373
T70 RUN_1 48.61 23.15 31.36 0.32617 0.12949
T70 RUN_2 70.00 11.95 20.42 0.28419 0.08731
T70 RUN_3 80.65 4.74 8.96 0.19270 0.03826
T70 RUN_4 31.22 36.43 33.63 0.32216 0.15688
T70 RUN_5 32.69 15.94 21.43 0.21717 0.05734
T81 RUN_1 4.54 66.03 8.50 0.11406 0.07716
T81 RUN_2 8.71 42.13 14.43 0.15560 0.06239
T81 RUN_3 13.51 28.46 18.33 0.17168 0.04657
T81 RUN_4 13.20 27.70 17.88 0.16667 0.05601
T81 RUN_5 21.35 22.20 21.77 0.20090 0.05283
T88 RUN_1 28.44 45.16 34.90 0.34146 0.20244
T88 RUN_2 28.17 45.92 34.92 0.34263 0.20069
T89 RUN_1 52.52 49.53 50.98 0.49997 0.28202
T89 RUN_2 52.02 48.96 50.44 0.49451 0.28589
T89 RUN_3 50.78 49.34 50.05 0.49016 0.27238
T89 RUN_4 52.50 49.91 51.17 0.50181 0.29220
T89 RUN_5 52.58 52.18 52.38 0.51382 0.29980
T89 SRVR_4 52.71 51.61 52.16 0.51163 0.29926
T89 SRVR_5 52.28 50.10 51.16 0.50168 0.30046
T89 SRVR_6 52.28 52.18 52.23 0.51226 0.30049
T89 SRVR_7 49.55 52.56 51.01 0.49972 0.29303
T89 SRVR_8 51.76 50.29 51.01 0.49999 0.29766
T90 RUN_1 53.33 47.06 50.00 0.49113 0.26805
T90 RUN_2 52.56 48.77 50.59 0.49625 0.28386
T90 RUN_3 52.30 58.25 0.5512 0.54201 0.35423
T90 RUN_4 61.09 38.14 46.96 0.47436 0.25209
T90 RUN_5 64.24 35.10 45.40 0.46707 0.24270
T100 RUN_1 44.59 51.61 47.85 0.46794 0.26055
T100 RUN_2 39.86 54.84 46.17 0.45448 0.26982
T100 RUN_3 35.29 44.59 39.40 0.38240 0.15734
T100 RUN_4 35.34 44.59 39.43 0.38271 0.15758
T100 RUN_5 54.86 18.22 27.35 0.30847 0.11109
  1. Micro-averaged results when evaluating all documents (i.e., measuring the overall performance of each run on the whole document set). The highest score for each evaluation column is show in bold typeface, the lowest in italics. Run/Srvr: RUN = offline run, SRVR = online server run via BCMS; MCC: Matthew's Correlation Coefficient; AUC iP/R: Area under the interpolated precision/recall curve (micro- averaged by iterating over the precision/recall values of the highest ranked annotation of all articles, then all second ranked annotations, etc.). Base_Top4: baseline system that assigns the four most frequent classes ordered by their frequencies in the training/development set. Base_Regex: simple matching strategy based on regular expressions.