TASS 2016. Official Results
As some groups have requested extra time for submission of non-official experiments, the annotated test sets will be released in a few days.
Task 1: Sentiment Analysis at global level (5 levels, Full test corpus)
| Group | Run Id | Date | Accuracy | Macro-P | Macro-R | Macro-F1 |
|---|---|---|---|---|---|---|
| jacerong | run-1-six-labels | 2016-08-18 17:17:45 | 0.614 | 0.471 | 0.531 | 0.499 |
| jacerong | run-2-six-labels | 2016-08-18 17:30:25 | 0.619 | 0.476 | 0.535 | 0.504 |
| jacerong | run-3-six-labels | 2016-08-18 17:39:11 | 0.620 | 0.477 | 0.532 | 0.503 |
| LABDA | RUN-1 | 2016-08-18 18:55:41 | 0.527 | 0.411 | 0.449 | 0.429 |
| LABDA | RUN-2 | 2016-08-18 18:59:51 | 0.527 | 0.412 | 0.448 | 0.429 |
| LABDA | RUN-3 | 2016-08-18 19:05:17 | 0.549 | 0.402 | 0.436 | 0.418 |
| INGEOTEC | B4MSA | 2016-08-18 19:39:54 | 0.535 | 0.429 | 0.505 | 0.464 |
| GASUCR | 01-noEmot-noPartNeg | 2016-08-23 04:52:18 | 0.225 | 0.339 | 0.177 | 0.232 |
| GASUCR | 01 | 2016-08-23 05:31:34 | 0.342 | 0.217 | 0.237 | 0.227 |
| GASUCR | 01-noEMO-noPartNeg | 2016-08-23 05:45:19 | 0.326 | 0.334 | 0.258 | 0.291 |
| ELiRF-UPV | ELiRF_task1-5levels_run1 | 2016-08-23 23:38:12 | 0.662 | 0.516 | 0.520 | 0.518 |
| ELiRF-UPV | ELiRF_task1-5levels_run2 | 2016-08-23 23:41:26 | 0.673 | 0.531 | 0.465 | 0.496 |
| GASUCR | 04 | 2016-08-24 19:47:57 | 0.410 | 0.268 | 0.242 | 0.254 |
(System ranking by accuracy)
Task 1: Sentiment Analysis at global level (3 levels, Full test corpus)
| Group | Run Id | Date | Accuracy | Macro-P | Macro-R | Macro-F1 |
|---|---|---|---|---|---|---|
| jacerong | run-1-four-labels | 2016-08-18 17:25:58 | 0.702 | 0.564 | 0.565 | 0.564 |
| jacerong | run-2-four-labels | 2016-08-18 17:35:14 | 0.704 | 0.567 | 0.568 | 0.567 |
| jacerong | run-3-four-labels | 2016-08-18 17:44:59 | 0.705 | 0.568 | 0.567 | 0.568 |
| LABDA | RUN-1 | 2016-08-18 19:10:00 | 0.652 | 0.506 | 0.510 | 0.508 |
| LABDA | RUN-2 | 2016-08-18 19:13:28 | 0.652 | 0.508 | 0.508 | 0.508 |
| LABDA | RUN-3 | 2016-08-18 19:16:47 | 0.653 | 0.512 | 0.511 | 0.511 |
| INGEOTEC | B4MSA | 2016-08-19 19:12:41 | 0.633 | 0.535 | 0.513 | 0.524 |
| GASUCR | 01-noEmot-noPartNeg | 2016-08-23 07:29:50 | 0.373 | 0.212 | 0.303 | 0.250 |
| GASUCR | 02-noEmot-noPartNeg | 2016-08-23 07:45:04 | 0.241 | 0.128 | 0.188 | 0.152 |
| ELiRF-UPV | ELiRF_task1-3levels_run1 | 2016-08-23 23:43:51 | 0.707 | 0.548 | 0.551 | 0.549 |
| ELiRF-UPV | ELiRF_task1-3levels_run2 | 2016-08-23 23:46:58 | 0.725 | 0.541 | 0.556 | 0.548 |
(System ranking by accuracy)
Task 2: Aspect based Sentiment Analysis (STOMPOL corpus)
| Group | Run Id | Date | Accuracy | Macro-P | Macro-R | Macro-F1 |
|---|---|---|---|---|---|---|
| GTI | GTI-Run1 | 2016-08-23 22:26:49 | 0.606 | 0.452 | 0.474 | 0.463 |
| ELiRF-UPV | ELiRF_task2_run1 | 2016-08-23 23:27:22 | 0.633 | 0.522 | 0.531 | 0.526 |
(System ranking by accuracy)
Gold Standard
- general-sentiment-5l.qrel : QREL file for General corpus test set, sentiment analysis, 5 levels
- general-sentiment-3l.qrel : QREL file for General corpus test set, sentiment analysis, 3 levels
- stompol-sentiment.qrel : QREL file for STOMPOL corpus test set, sentiment analysis
