Datasets:
Remove conceptMention id and unify labels, update README.md
Browse files
README.md
CHANGED
@@ -196,16 +196,16 @@ dataset_info:
|
|
196 |
'40': I-trigger
|
197 |
splits:
|
198 |
- name: train
|
199 |
-
num_bytes:
|
200 |
num_examples: 788
|
201 |
- name: test
|
202 |
-
num_bytes:
|
203 |
num_examples: 484
|
204 |
- name: validation
|
205 |
-
num_bytes:
|
206 |
num_examples: 152
|
207 |
download_size: 8190212
|
208 |
-
dataset_size:
|
209 |
- config_name: el
|
210 |
features:
|
211 |
- name: id
|
@@ -436,9 +436,9 @@ dataset_info:
|
|
436 |
- **Repository:** [https://github.com/dfki-nlp/mobie](https://github.com/dfki-nlp/mobie)
|
437 |
- **Paper:** [https://aclanthology.org/2021.konvens-1.22/](https://aclanthology.org/2021.konvens-1.22/)
|
438 |
- **Point of Contact:** See [https://github.com/dfki-nlp/mobie](https://github.com/dfki-nlp/mobie)
|
439 |
-
- **Size of downloaded dataset files:**
|
440 |
-
- **Size of the generated dataset:** 1.
|
441 |
-
- **Total amount of disk used:** 9.
|
442 |
|
443 |
### Dataset Summary
|
444 |
|
@@ -446,13 +446,18 @@ This script is for loading the MobIE dataset from https://github.com/dfki-nlp/mo
|
|
446 |
|
447 |
MobIE is a German-language dataset which is human-annotated with 20 coarse- and fine-grained entity types and entity linking information for geographically linkable entities. The dataset consists of 3,232 social media texts and traffic reports with 91K tokens, and contains 20.5K annotated entities, 13.1K of which are linked to a knowledge base. A subset of the dataset is human-annotated with seven mobility-related, n-ary relation types, while the remaining documents are annotated using a weakly-supervised labeling approach implemented with the Snorkel framework. The dataset combines annotations for NER, EL and RE, and thus can be used for joint and multi-task learning of these fundamental information extraction tasks.
|
448 |
|
449 |
-
This version of the dataset loader provides
|
|
|
|
|
|
|
|
|
|
|
450 |
|
451 |
For more details see https://github.com/dfki-nlp/mobie and https://aclanthology.org/2021.konvens-1.22/.
|
452 |
|
453 |
### Supported Tasks and Leaderboards
|
454 |
|
455 |
-
- **Tasks:** Named Entity Recognition
|
456 |
- **Leaderboards:**
|
457 |
|
458 |
### Languages
|
@@ -463,34 +468,494 @@ German
|
|
463 |
|
464 |
### Data Instances
|
465 |
|
466 |
-
|
467 |
-
- **Size of
|
468 |
-
- **
|
|
|
469 |
|
470 |
An example of 'train' looks as follows.
|
471 |
|
472 |
```json
|
473 |
{
|
474 |
-
|
475 |
-
|
476 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
477 |
}
|
478 |
```
|
479 |
|
480 |
|
481 |
### Data Fields
|
482 |
|
483 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
484 |
|
485 |
-
- `id`: a `string` feature.
|
486 |
-
- `
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
487 |
- `ner_tags`: a `list` of classification labels, with possible values including `O` (0), `B-date` (1), `I-date` (2), `B-disaster-type` (3), `I-disaster-type` (4), ...
|
488 |
|
489 |
### Data Splits
|
490 |
|
491 |
-
|
|
492 |
-
|
493 |
-
|
|
|
|
|
|
|
|
494 |
|
495 |
## Dataset Creation
|
496 |
|
|
|
196 |
'40': I-trigger
|
197 |
splits:
|
198 |
- name: train
|
199 |
+
num_bytes: 1869427
|
200 |
num_examples: 788
|
201 |
- name: test
|
202 |
+
num_bytes: 1117030
|
203 |
num_examples: 484
|
204 |
- name: validation
|
205 |
+
num_bytes: 365928
|
206 |
num_examples: 152
|
207 |
download_size: 8190212
|
208 |
+
dataset_size: 3352385
|
209 |
- config_name: el
|
210 |
features:
|
211 |
- name: id
|
|
|
436 |
- **Repository:** [https://github.com/dfki-nlp/mobie](https://github.com/dfki-nlp/mobie)
|
437 |
- **Paper:** [https://aclanthology.org/2021.konvens-1.22/](https://aclanthology.org/2021.konvens-1.22/)
|
438 |
- **Point of Contact:** See [https://github.com/dfki-nlp/mobie](https://github.com/dfki-nlp/mobie)
|
439 |
+
- **Size of downloaded dataset files:** 8.2 MB
|
440 |
+
- **Size of the generated dataset:** 1.7 MB
|
441 |
+
- **Total amount of disk used:** 9.9 MB
|
442 |
|
443 |
### Dataset Summary
|
444 |
|
|
|
446 |
|
447 |
MobIE is a German-language dataset which is human-annotated with 20 coarse- and fine-grained entity types and entity linking information for geographically linkable entities. The dataset consists of 3,232 social media texts and traffic reports with 91K tokens, and contains 20.5K annotated entities, 13.1K of which are linked to a knowledge base. A subset of the dataset is human-annotated with seven mobility-related, n-ary relation types, while the remaining documents are annotated using a weakly-supervised labeling approach implemented with the Snorkel framework. The dataset combines annotations for NER, EL and RE, and thus can be used for joint and multi-task learning of these fundamental information extraction tasks.
|
448 |
|
449 |
+
This version of the dataset loader provides configurations for:
|
450 |
+
|
451 |
+
- Named Entity Recognition (`ner`): NER tags use the `BIO` tagging scheme
|
452 |
+
- Entity Linking (`el`): Entity mentions are linked to an internal knowledge base and Open Street Map
|
453 |
+
- Relation Extraction (`re`): n-ary Relation Extraction
|
454 |
+
- Event Extraction (`ee`): formatted similar to https://github.com/nlpcl-lab/ace2005-preprocessing?tab=readme-ov-file#format
|
455 |
|
456 |
For more details see https://github.com/dfki-nlp/mobie and https://aclanthology.org/2021.konvens-1.22/.
|
457 |
|
458 |
### Supported Tasks and Leaderboards
|
459 |
|
460 |
+
- **Tasks:** Named Entity Recognition, Entity Linking, n-ary Relation Extraction, Event Extraction
|
461 |
- **Leaderboards:**
|
462 |
|
463 |
### Languages
|
|
|
468 |
|
469 |
### Data Instances
|
470 |
|
471 |
+
#### ner
|
472 |
+
- **Size of downloaded dataset files:** 8.2 MB
|
473 |
+
- **Size of the generated dataset:** 1.7 MB
|
474 |
+
- **Total amount of disk used:** 10.9 MB
|
475 |
|
476 |
An example of 'train' looks as follows.
|
477 |
|
478 |
```json
|
479 |
{
|
480 |
+
"id": "http://www.ndr.de/nachrichten/verkehr/index.html#2@2016-05-04T21:02:14.000+02:00",
|
481 |
+
"tokens": ["Vorsicht", "bitte", "auf", "der", "A28", "Leer", "Richtung", "Oldenburg", "zwischen", "Zwischenahner", "Meer", "und", "Neuenkruge", "liegen", "Gegenstände", "!"],
|
482 |
+
"ner_tags": [0, 0, 0, 0, 19, 13, 0, 13, 0, 11, 12, 0, 11, 0, 0, 0]
|
483 |
+
}
|
484 |
+
```
|
485 |
+
|
486 |
+
#### el
|
487 |
+
- **Size of downloaded dataset files:** 8.2 MB
|
488 |
+
- **Size of the generated dataset:** 2.1 MB
|
489 |
+
- **Total amount of disk used:** 10.3 MB
|
490 |
+
|
491 |
+
An example of 'train' looks as follows.
|
492 |
+
|
493 |
+
```json
|
494 |
+
{
|
495 |
+
"id": "1108129826844672001",
|
496 |
+
"text": "#S4 #RegioNDS #Teilausfall #Mellendorf(23.03)> #Bennemühlen(23.07). Grund: technische Störung an der Strecke. Bitte nutzen Sie #RB38 nach Soltau über Bennemühlen Abfahrt: 23:08 Uhr vom Gleis 2",
|
497 |
+
"entity_mentions": [
|
498 |
+
{
|
499 |
+
"text": "#S4",
|
500 |
+
"start": 0,
|
501 |
+
"end": 1,
|
502 |
+
"char_start": 0,
|
503 |
+
"char_end": 3,
|
504 |
+
"type": 7,
|
505 |
+
"entity_id": "NIL",
|
506 |
+
"refids": [
|
507 |
+
{
|
508 |
+
"key": "spreeDBReferenceId",
|
509 |
+
"value": "24007"
|
510 |
+
}
|
511 |
+
]
|
512 |
+
},
|
513 |
+
{
|
514 |
+
"text": "#RegioNDS",
|
515 |
+
"start": 1,
|
516 |
+
"end": 2,
|
517 |
+
"char_start": 4,
|
518 |
+
"char_end": 13,
|
519 |
+
"type": 13,
|
520 |
+
"entity_id": "NIL",
|
521 |
+
"refids": [
|
522 |
+
{
|
523 |
+
"key": "spreeDBReferenceId",
|
524 |
+
"value": "NIL"
|
525 |
+
}
|
526 |
+
]
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"text": "#Teilausfall",
|
530 |
+
"start": 2,
|
531 |
+
"end": 3,
|
532 |
+
"char_start": 14,
|
533 |
+
"char_end": 26,
|
534 |
+
"type": 19,
|
535 |
+
"entity_id": "NIL",
|
536 |
+
"refids": [
|
537 |
+
{
|
538 |
+
"key": "spreeDBReferenceId",
|
539 |
+
"value": "NIL"
|
540 |
+
}
|
541 |
+
]
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"text": "#Mellendorf",
|
545 |
+
"start": 3,
|
546 |
+
"end": 4,
|
547 |
+
"char_start": 27,
|
548 |
+
"char_end": 38,
|
549 |
+
"type": 8,
|
550 |
+
"entity_id": "NIL",
|
551 |
+
"refids": [
|
552 |
+
{
|
553 |
+
"key": "spreeDBReferenceId",
|
554 |
+
"value": "8003957"
|
555 |
+
}
|
556 |
+
]
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"text": "23.03",
|
560 |
+
"start": 5,
|
561 |
+
"end": 6,
|
562 |
+
"char_start": 39,
|
563 |
+
"char_end": 44,
|
564 |
+
"type": 0,
|
565 |
+
"entity_id": "NIL",
|
566 |
+
"refids": [
|
567 |
+
{
|
568 |
+
"key": "spreeDBReferenceId",
|
569 |
+
"value": "NIL"
|
570 |
+
}
|
571 |
+
]
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"text": "#Bennemühlen",
|
575 |
+
"start": 8,
|
576 |
+
"end": 9,
|
577 |
+
"char_start": 47,
|
578 |
+
"char_end": 59,
|
579 |
+
"type": 6,
|
580 |
+
"entity_id": "29589800",
|
581 |
+
"refids": [
|
582 |
+
{
|
583 |
+
"key": "spreeDBReferenceId",
|
584 |
+
"value": "29589800"
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"key": "osm_id",
|
588 |
+
"value": "29589800"
|
589 |
+
}
|
590 |
+
]
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"text": "23.07",
|
594 |
+
"start": 10,
|
595 |
+
"end": 11,
|
596 |
+
"char_start": 60,
|
597 |
+
"char_end": 65,
|
598 |
+
"type": 0,
|
599 |
+
"entity_id": "NIL",
|
600 |
+
"refids": [
|
601 |
+
{
|
602 |
+
"key": "spreeDBReferenceId",
|
603 |
+
"value": "NIL"
|
604 |
+
}
|
605 |
+
]
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"text": "technische Störung",
|
609 |
+
"start": 15,
|
610 |
+
"end": 17,
|
611 |
+
"char_start": 76,
|
612 |
+
"char_end": 94,
|
613 |
+
"type": 4,
|
614 |
+
"entity_id": "NIL",
|
615 |
+
"refids": [
|
616 |
+
{
|
617 |
+
"key": "spreeDBReferenceId",
|
618 |
+
"value": "NIL"
|
619 |
+
}
|
620 |
+
]
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"text": "#RB38",
|
624 |
+
"start": 24,
|
625 |
+
"end": 25,
|
626 |
+
"char_start": 128,
|
627 |
+
"char_end": 133,
|
628 |
+
"type": 7,
|
629 |
+
"entity_id": "NIL",
|
630 |
+
"refids": [
|
631 |
+
{
|
632 |
+
"key": "spreeDBReferenceId",
|
633 |
+
"value": "23138"
|
634 |
+
}
|
635 |
+
]
|
636 |
+
},
|
637 |
+
{
|
638 |
+
"text": "Soltau",
|
639 |
+
"start": 26,
|
640 |
+
"end": 27,
|
641 |
+
"char_start": 139,
|
642 |
+
"char_end": 145,
|
643 |
+
"type": 6,
|
644 |
+
"entity_id": "1809016",
|
645 |
+
"refids": [
|
646 |
+
{
|
647 |
+
"key": "spreeDBReferenceId",
|
648 |
+
"value": "-1809016"
|
649 |
+
},
|
650 |
+
{
|
651 |
+
"key": "osm_id",
|
652 |
+
"value": "1809016"
|
653 |
+
}
|
654 |
+
]
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"text": "Bennemühlen",
|
658 |
+
"start": 28,
|
659 |
+
"end": 29,
|
660 |
+
"char_start": 151,
|
661 |
+
"char_end": 162,
|
662 |
+
"type": 8,
|
663 |
+
"entity_id": "NIL",
|
664 |
+
"refids": [
|
665 |
+
{
|
666 |
+
"key": "spreeDBReferenceId",
|
667 |
+
"value": "8000871"
|
668 |
+
}
|
669 |
+
]
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"text": "23:08 Uhr",
|
673 |
+
"start": 31,
|
674 |
+
"end": 33,
|
675 |
+
"char_start": 172,
|
676 |
+
"char_end": 181,
|
677 |
+
"type": 18,
|
678 |
+
"entity_id": "NIL",
|
679 |
+
"refids": [
|
680 |
+
{
|
681 |
+
"key": "spreeDBReferenceId",
|
682 |
+
"value": "NIL"
|
683 |
+
}
|
684 |
+
]
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"text": "2",
|
688 |
+
"start": 35,
|
689 |
+
"end": 36,
|
690 |
+
"char_start": 192,
|
691 |
+
"char_end": 193,
|
692 |
+
"type": 11,
|
693 |
+
"entity_id": "NIL",
|
694 |
+
"refids": [
|
695 |
+
{
|
696 |
+
"key": "spreeDBReferenceId",
|
697 |
+
"value": "NIL"
|
698 |
+
}
|
699 |
+
]
|
700 |
+
}
|
701 |
+
]
|
702 |
+
}
|
703 |
+
```
|
704 |
+
|
705 |
+
#### re
|
706 |
+
- **Size of downloaded dataset files:** 8.2 MB
|
707 |
+
- **Size of the generated dataset:** 1.7 MB
|
708 |
+
- **Total amount of disk used:** 10.9 MB
|
709 |
+
|
710 |
+
An example of 'train' looks as follows.
|
711 |
+
|
712 |
+
```json
|
713 |
+
{
|
714 |
+
"id": "1111185208647274501_1",
|
715 |
+
"text": "RT @SBahn_Stuttgart: 🚨Störung🚨 Derzeit steht eine #S2 Richtung Filderstadt mit einer Türstörung in Stg-Rohr. Es kommt auf den Linien #S1, #…",
|
716 |
+
"tokens": ["RT", "@SBahn_Stuttgart", ":", "🚨", "Störung", "🚨 ", "Derzeit", "steht", "eine", "#S2", "Richtung", "Filderstadt", "mit", "einer", "Türstörung", "in", "Stg", "-", "Rohr", ".", "Es", "kommt", "auf", "den", "Linien", "#S1", ",", "#", "…"],
|
717 |
+
"entities": [[1, 2], [4, 5], [9, 10], [11, 12], [14, 15], [16, 19], [25, 26]],
|
718 |
+
"entity_roles": [0, 1, 2, 0, 0, 0, 0],
|
719 |
+
"entity_types": [13, 4, 7, 6, 4, 8, 7],
|
720 |
+
"event_type": 5,
|
721 |
+
"entity_ids": ["NIL", "NIL", "NIL", "2796535", "NIL", "NIL", "NIL"]
|
722 |
+
}
|
723 |
+
```
|
724 |
+
|
725 |
+
#### ee
|
726 |
+
- **Size of downloaded dataset files:** 8.2 MB
|
727 |
+
- **Size of the generated dataset:** 3.7 MB
|
728 |
+
- **Total amount of disk used:** 11.9 MB
|
729 |
+
|
730 |
+
An example of 'train' looks as follows.
|
731 |
+
|
732 |
+
```json
|
733 |
+
{
|
734 |
+
"id": "1111185208647274501",
|
735 |
+
"text": "RT @SBahn_Stuttgart: 🚨Störung🚨 Derzeit steht eine #S2 Richtung Filderstadt mit einer Türstörung in Stg-Rohr. Es kommt auf den Linien #S1, #…",
|
736 |
+
"entity_mentions": [
|
737 |
+
{
|
738 |
+
"text": "@SBahn_Stuttgart",
|
739 |
+
"start": 1,
|
740 |
+
"end": 2,
|
741 |
+
"char_start": 3,
|
742 |
+
"char_end": 19,
|
743 |
+
"type": 13,
|
744 |
+
"entity_id": "NIL",
|
745 |
+
"refids": [
|
746 |
+
{
|
747 |
+
"key": "spreeDBReferenceId",
|
748 |
+
"value": "NIL"
|
749 |
+
}
|
750 |
+
]
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"text": "Störung",
|
754 |
+
"start": 4,
|
755 |
+
"end": 5,
|
756 |
+
"char_start": 22,
|
757 |
+
"char_end": 29,
|
758 |
+
"type": 4,
|
759 |
+
"entity_id": "NIL",
|
760 |
+
"refids": [
|
761 |
+
{
|
762 |
+
"key": "spreeDBReferenceId",
|
763 |
+
"value": "NIL"
|
764 |
+
}
|
765 |
+
]
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"text": "#S2",
|
769 |
+
"start": 9,
|
770 |
+
"end": 10,
|
771 |
+
"char_start": 50,
|
772 |
+
"char_end": 53,
|
773 |
+
"type": 7,
|
774 |
+
"entity_id": "NIL",
|
775 |
+
"refids": [
|
776 |
+
{
|
777 |
+
"key": "spreeDBReferenceId",
|
778 |
+
"value": "17171"
|
779 |
+
}
|
780 |
+
]
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"text": "Filderstadt",
|
784 |
+
"start": 11,
|
785 |
+
"end": 12,
|
786 |
+
"char_start": 63,
|
787 |
+
"char_end": 74,
|
788 |
+
"type": 6,
|
789 |
+
"entity_id": "2796535",
|
790 |
+
"refids": [
|
791 |
+
{
|
792 |
+
"key": "spreeDBReferenceId",
|
793 |
+
"value": "-2796535"
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"key": "osm_id",
|
797 |
+
"value": "2796535"
|
798 |
+
}
|
799 |
+
]
|
800 |
+
},
|
801 |
+
{
|
802 |
+
"text": "Türstörung",
|
803 |
+
"start": 14,
|
804 |
+
"end": 15,
|
805 |
+
"char_start": 85,
|
806 |
+
"char_end": 95,
|
807 |
+
"type": 4,
|
808 |
+
"entity_id": "NIL",
|
809 |
+
"refids": [
|
810 |
+
{
|
811 |
+
"key": "spreeDBReferenceId",
|
812 |
+
"value": "NIL"
|
813 |
+
}
|
814 |
+
]
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"text": "Stg-Rohr",
|
818 |
+
"start": 16,
|
819 |
+
"end": 19,
|
820 |
+
"char_start": 99,
|
821 |
+
"char_end": 107,
|
822 |
+
"type": 8,
|
823 |
+
"entity_id": "NIL",
|
824 |
+
"refids": [
|
825 |
+
{
|
826 |
+
"key": "spreeDBReferenceId",
|
827 |
+
"value": "NIL"
|
828 |
+
}
|
829 |
+
]
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"text": "#S1",
|
833 |
+
"start": 25,
|
834 |
+
"end": 26,
|
835 |
+
"char_start": 133,
|
836 |
+
"char_end": 136,
|
837 |
+
"type": 7,
|
838 |
+
"entity_id": "NIL",
|
839 |
+
"refids": [
|
840 |
+
{
|
841 |
+
"key": "spreeDBReferenceId",
|
842 |
+
"value": "16703"
|
843 |
+
}
|
844 |
+
]
|
845 |
+
}
|
846 |
+
],
|
847 |
+
"event_mentions": [
|
848 |
+
{
|
849 |
+
"id": "r/0f748b57-63ec-4cb9-ab54-e35d29ac44f8",
|
850 |
+
"trigger": {
|
851 |
+
"text": "Störung",
|
852 |
+
"start": 4,
|
853 |
+
"end": 5,
|
854 |
+
"char_start": 22,
|
855 |
+
"char_end": 29
|
856 |
+
},
|
857 |
+
"arguments": [
|
858 |
+
{
|
859 |
+
"text": "#S2",
|
860 |
+
"start": 9,
|
861 |
+
"end": 10,
|
862 |
+
"char_start": 50,
|
863 |
+
"char_end": 53,
|
864 |
+
"role": 1,
|
865 |
+
"type": 7
|
866 |
+
}
|
867 |
+
],
|
868 |
+
"event_type": 5
|
869 |
+
}
|
870 |
+
],
|
871 |
+
"tokens": ["RT", "@SBahn_Stuttgart", ":", "🚨", "Störung", "🚨 ", "Derzeit", "steht", "eine", "#S2", "Richtung", "Filderstadt", "mit", "einer", "Türstörung", "in", "Stg", "-", "Rohr", ".", "Es", "kommt", "auf", "den", "Linien", "#S1", ",", "#", "…"],
|
872 |
+
"pos_tags": ["NN", "NN", "$.", "CARD", "NN", "CARD", "ADV", "VVFIN", "ART", "NN", "NN", "NE", "APPR", "ART", "NN", "APPR", "NE", "$[", "NE", "$.", "PPER", "VVFIN", "APPR", "ART", "NN", "CARD", "$,", "CARD", "$["],
|
873 |
+
"lemma": ["rt", "@sbahn_stuttgart", ":", "🚨", "störung", "🚨", "derzeit", "steht", "eine", "#s2", "richtung", "filderstadt", "mit", "einer", "türstörung", "in", "stg", "-", "rohr", ".", "es", "kommt", "auf", "den", "linien", "#s1", ",", "#", "..."],
|
874 |
+
"ner_tags": [0, 14, 0, 0, 5, 0, 0, 0, 0, 8, 0, 7, 0, 0, 5, 0, 9, 29, 29, 0, 0, 0, 0, 0, 0, 8, 0, 0, 0]
|
875 |
}
|
876 |
```
|
877 |
|
878 |
|
879 |
### Data Fields
|
880 |
|
881 |
+
#### ner
|
882 |
+
|
883 |
+
- `id`: example identifier, a `string` feature.
|
884 |
+
- `tokens`: list of tokens, a `list` of `string` features.
|
885 |
+
- `ner_tags`: a `list` of classification labels, with possible values including `O` (0), `B-date` (1), `I-date` (2), `B-disaster-type` (3), `I-disaster-type` (4), ...
|
886 |
+
|
887 |
+
#### el
|
888 |
+
|
889 |
+
- `id`: example identifier, a `string` feature.
|
890 |
+
- `text`: example text, a `string` feature.
|
891 |
+
- `entity_mentions`: a `list` of `struct` features.
|
892 |
+
- `text`: a `string` feature.
|
893 |
+
- `start`: token offset start, a `int32` feature.
|
894 |
+
- `end`: token offset end, a `int32` feature.
|
895 |
+
- `char_start`: character offset start, a `int32` feature.
|
896 |
+
- `char_end`: character offset end, a `int32` feature.
|
897 |
+
- `type`: a classification label, with possible values including `O` (0), `date` (1), `disaster-type` (2), `distance` (3), `duration` (4), `event-cause` (5), ...
|
898 |
+
- `entity_id`: Open Street Map ID, a `string` feature.
|
899 |
+
- `refids`: knowledge base ids, a `list` of `struct` features.
|
900 |
+
- `key`: name of the knowledge base, a `string` feature.
|
901 |
+
- `value`: identifier, a `string` feature.
|
902 |
+
|
903 |
+
|
904 |
+
#### re
|
905 |
+
|
906 |
+
- `id`: example identifier, a `string` feature.
|
907 |
+
- `text`: example text, a `string` feature.
|
908 |
+
- `tokens`: list of tokens, a `list` of `string` features.
|
909 |
+
- `entities`: a list of token spans, a `list` of `int32` featuress.
|
910 |
+
- `entity_roles`: a `list` of classification labels, with possible values including `no_arg` (0), `trigger` (1), `location` (2), `delay` (3), `direction` (4), ...
|
911 |
+
- `event_type`: a classification label, with possible values including `O` (0), `Accident` (1), `CanceledRoute` (2), `CanceledStop` (3), `Delay` (4), ...
|
912 |
+
- `entity_ids`: list of Open Street Map IDs, a `list` of `string` features.
|
913 |
+
|
914 |
+
#### ee
|
915 |
|
916 |
+
- `id`: example identifier, a `string` feature.
|
917 |
+
- `text`: example text, a `string` feature.
|
918 |
+
- `entity_mentions`: a `list` of `struct` features.
|
919 |
+
- `text`: a `string` feature.
|
920 |
+
- `start`: token offset start, a `int32` feature.
|
921 |
+
- `end`: token offset end, a `int32` feature.
|
922 |
+
- `char_start`: character offset start, a `int32` feature.
|
923 |
+
- `char_end`: character offset end, a `int32` feature.
|
924 |
+
- `type`: a classification label, with possible values including `O` (0), `date` (1), `disaster-type` (2), `distance` (3), `duration` (4), `event-cause` (5), ...
|
925 |
+
- `entity_id`: Open Street Map ID, a `string` feature.
|
926 |
+
- `refids`: knowledge base ids, a `list` of `struct` features.
|
927 |
+
- `key`: name of the knowledge base, a `string` feature.
|
928 |
+
- `value`: identifier, a `string` feature.
|
929 |
+
- `event_mentions`: a list of `struct` features.
|
930 |
+
- `id`: event identifier, a `string` feature.
|
931 |
+
- `trigger`: a `struct` feature.
|
932 |
+
- `text`: a `string` feature.
|
933 |
+
- `start`: token offset start, a `int32` feature.
|
934 |
+
- `end`: token offset end, a `int32` feature.
|
935 |
+
- `char_start`: character offset start, a `int32` feature.
|
936 |
+
- `char_end`: character offset end, a `int32` feature.
|
937 |
+
- `arguments`: a list of `struct` features.
|
938 |
+
- `text`: a `string` feature.
|
939 |
+
- `start`: token offset start, a `int32` feature.
|
940 |
+
- `end`: token offset end, a `int32` feature.
|
941 |
+
- `char_start`: character offset start, a `int32` feature.
|
942 |
+
- `char_end`: character offset end, a `int32` feature.
|
943 |
+
- `role`: a classification label, with possible values including `no_arg` (0), `trigger` (1), `location` (2), `delay` (3), `direction` (4), ...
|
944 |
+
- `type`: a classification label, with possible values including `O` (0), `date` (1), `disaster-type` (2), `distance` (3), `duration` (4), `event-cause` (5), ...
|
945 |
+
- `event_type`: a classification label, with possible values including `O` (0), `Accident` (1), `CanceledRoute` (2), `CanceledStop` (3), `Delay` (4), ...
|
946 |
+
- `tokens`: list of tokens, a `list` of `string` features.
|
947 |
+
- `pos_tags`: list of part-of-speech tags, a `list` of `string` features.
|
948 |
+
- `lemma`: list of lemmatized tokens, a `list` of `string` features.
|
949 |
- `ner_tags`: a `list` of classification labels, with possible values including `O` (0), `B-date` (1), `I-date` (2), `B-disaster-type` (3), `I-disaster-type` (4), ...
|
950 |
|
951 |
### Data Splits
|
952 |
|
953 |
+
| | Train | Dev | Test |
|
954 |
+
|-----|-------|-----|------|
|
955 |
+
| NER | 2115 | 494 | 623 |
|
956 |
+
| EL | 2115 | 494 | 623 |
|
957 |
+
| RE | 1199 | 228 | 609 |
|
958 |
+
| EE | 788 | 152 | 484 |
|
959 |
|
960 |
## Dataset Creation
|
961 |
|
mobie.py
CHANGED
@@ -136,7 +136,7 @@ def fix_doc(doc):
|
|
136 |
class Mobie(datasets.GeneratorBasedBuilder):
|
137 |
"""MobIE is a German-language dataset which is human-annotated with 20 coarse- and fine-grained entity types and entity linking information for geographically linkable entities"""
|
138 |
|
139 |
-
VERSION = datasets.Version("1.
|
140 |
|
141 |
# This is an example of a dataset with multiple configurations.
|
142 |
# If you don't want/need to define several sub-sets in your dataset,
|
@@ -180,13 +180,23 @@ class Mobie(datasets.GeneratorBasedBuilder):
|
|
180 |
"time",
|
181 |
"trigger",
|
182 |
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
183 |
entity_mentions = [
|
184 |
{
|
185 |
-
"id": datasets.Value("string"),
|
186 |
"text": datasets.Value("string"),
|
187 |
"start": datasets.Value("int32"),
|
188 |
"end": datasets.Value("int32"),
|
189 |
-
"
|
|
|
|
|
|
|
190 |
"refids": [
|
191 |
{
|
192 |
"key": datasets.Value("string"),
|
@@ -224,19 +234,13 @@ class Mobie(datasets.GeneratorBasedBuilder):
|
|
224 |
"tokens": datasets.Sequence(datasets.Value("string")),
|
225 |
"entities": datasets.Sequence([datasets.Value("int32")]),
|
226 |
"entity_roles": datasets.Sequence(datasets.features.ClassLabel(
|
227 |
-
names=
|
228 |
-
"no_arg", "trigger", "location", "delay", "direction", "start_loc", "end_loc",
|
229 |
-
"start_date", "end_date", "cause", "jam_length", "route"
|
230 |
-
]
|
231 |
)),
|
232 |
"entity_types": datasets.Sequence(datasets.features.ClassLabel(
|
233 |
-
names=labels
|
234 |
)),
|
235 |
"event_type": datasets.features.ClassLabel(
|
236 |
-
names=
|
237 |
-
"O", "Accident", "CanceledRoute", "CanceledStop", "Delay", "Obstruction",
|
238 |
-
"RailReplacementService", "TrafficJam"
|
239 |
-
]
|
240 |
),
|
241 |
"entity_ids": datasets.Sequence(datasets.Value("string"))
|
242 |
}
|
@@ -252,31 +256,27 @@ class Mobie(datasets.GeneratorBasedBuilder):
|
|
252 |
{
|
253 |
"id": datasets.Value("string"),
|
254 |
"trigger": {
|
255 |
-
"id": datasets.Value("string"),
|
256 |
"text": datasets.Value("string"),
|
257 |
"start": datasets.Value("int32"),
|
258 |
"end": datasets.Value("int32"),
|
|
|
|
|
259 |
},
|
260 |
"arguments": [{
|
261 |
-
"id": datasets.Value("string"),
|
262 |
"text": datasets.Value("string"),
|
263 |
"start": datasets.Value("int32"),
|
264 |
"end": datasets.Value("int32"),
|
|
|
|
|
265 |
"role": datasets.features.ClassLabel(
|
266 |
-
names=
|
267 |
-
"no_arg", "location", "delay", "direction", "start_loc", "end_loc",
|
268 |
-
"start_date", "end_date", "cause", "jam_length", "route"
|
269 |
-
]
|
270 |
),
|
271 |
"type": datasets.features.ClassLabel(
|
272 |
-
names=labels
|
273 |
)
|
274 |
}],
|
275 |
"event_type": datasets.features.ClassLabel(
|
276 |
-
names=
|
277 |
-
"O", "Accident", "CanceledRoute", "CanceledStop", "Delay", "Obstruction",
|
278 |
-
"RailReplacementService", "TrafficJam"
|
279 |
-
]
|
280 |
),
|
281 |
}
|
282 |
],
|
@@ -497,7 +497,6 @@ class Mobie(datasets.GeneratorBasedBuilder):
|
|
497 |
assert arg_start != -1 and arg_end != -1, f"Could not find token offsets for {arg['conceptMention']['id']}"
|
498 |
arg_text = text[arg_char_start:arg_char_end]
|
499 |
args.append({
|
500 |
-
"id": arg["conceptMention"]["id"],
|
501 |
"text": arg_text,
|
502 |
"start": arg_start,
|
503 |
"end": arg_end,
|
@@ -509,7 +508,6 @@ class Mobie(datasets.GeneratorBasedBuilder):
|
|
509 |
event_mentions.append({
|
510 |
"id": rm["id"],
|
511 |
"trigger": {
|
512 |
-
"id": trigger["conceptMention"]["id"],
|
513 |
"text": trigger_text,
|
514 |
"start": trigger_start,
|
515 |
"end": trigger_end,
|
|
|
136 |
class Mobie(datasets.GeneratorBasedBuilder):
|
137 |
"""MobIE is a German-language dataset which is human-annotated with 20 coarse- and fine-grained entity types and entity linking information for geographically linkable entities"""
|
138 |
|
139 |
+
VERSION = datasets.Version("1.1.0")
|
140 |
|
141 |
# This is an example of a dataset with multiple configurations.
|
142 |
# If you don't want/need to define several sub-sets in your dataset,
|
|
|
180 |
"time",
|
181 |
"trigger",
|
182 |
]
|
183 |
+
event_types = [
|
184 |
+
"O", "Accident", "CanceledRoute", "CanceledStop", "Delay", "Obstruction",
|
185 |
+
"RailReplacementService", "TrafficJam"
|
186 |
+
]
|
187 |
+
event_roles = [
|
188 |
+
"no_arg", "trigger", "location", "delay", "direction", "start_loc", "end_loc",
|
189 |
+
"start_date", "end_date", "cause", "jam_length", "route"
|
190 |
+
]
|
191 |
entity_mentions = [
|
192 |
{
|
|
|
193 |
"text": datasets.Value("string"),
|
194 |
"start": datasets.Value("int32"),
|
195 |
"end": datasets.Value("int32"),
|
196 |
+
"char_start": datasets.Value("int32"),
|
197 |
+
"char_end": datasets.Value("int32"),
|
198 |
+
"type": datasets.features.ClassLabel(names=["O"]+labels),
|
199 |
+
"entity_id": datasets.Value("string"),
|
200 |
"refids": [
|
201 |
{
|
202 |
"key": datasets.Value("string"),
|
|
|
234 |
"tokens": datasets.Sequence(datasets.Value("string")),
|
235 |
"entities": datasets.Sequence([datasets.Value("int32")]),
|
236 |
"entity_roles": datasets.Sequence(datasets.features.ClassLabel(
|
237 |
+
names=event_roles
|
|
|
|
|
|
|
238 |
)),
|
239 |
"entity_types": datasets.Sequence(datasets.features.ClassLabel(
|
240 |
+
names=["O"] + labels
|
241 |
)),
|
242 |
"event_type": datasets.features.ClassLabel(
|
243 |
+
names=event_types
|
|
|
|
|
|
|
244 |
),
|
245 |
"entity_ids": datasets.Sequence(datasets.Value("string"))
|
246 |
}
|
|
|
256 |
{
|
257 |
"id": datasets.Value("string"),
|
258 |
"trigger": {
|
|
|
259 |
"text": datasets.Value("string"),
|
260 |
"start": datasets.Value("int32"),
|
261 |
"end": datasets.Value("int32"),
|
262 |
+
"char_start": datasets.Value("int32"),
|
263 |
+
"char_end": datasets.Value("int32")
|
264 |
},
|
265 |
"arguments": [{
|
|
|
266 |
"text": datasets.Value("string"),
|
267 |
"start": datasets.Value("int32"),
|
268 |
"end": datasets.Value("int32"),
|
269 |
+
"char_start": datasets.Value("int32"),
|
270 |
+
"char_end": datasets.Value("int32"),
|
271 |
"role": datasets.features.ClassLabel(
|
272 |
+
names=event_roles
|
|
|
|
|
|
|
273 |
),
|
274 |
"type": datasets.features.ClassLabel(
|
275 |
+
names=["O"] + labels
|
276 |
)
|
277 |
}],
|
278 |
"event_type": datasets.features.ClassLabel(
|
279 |
+
names=event_types
|
|
|
|
|
|
|
280 |
),
|
281 |
}
|
282 |
],
|
|
|
497 |
assert arg_start != -1 and arg_end != -1, f"Could not find token offsets for {arg['conceptMention']['id']}"
|
498 |
arg_text = text[arg_char_start:arg_char_end]
|
499 |
args.append({
|
|
|
500 |
"text": arg_text,
|
501 |
"start": arg_start,
|
502 |
"end": arg_end,
|
|
|
508 |
event_mentions.append({
|
509 |
"id": rm["id"],
|
510 |
"trigger": {
|
|
|
511 |
"text": trigger_text,
|
512 |
"start": trigger_start,
|
513 |
"end": trigger_end,
|