Dataset Summary
A dataset for benchmarking keyphrase extraction and generation techniques from english news articles. For more details about the dataset please refer the original paper - https://arxiv.org/abs/1704.02853
Original source of the data - https://github.com/LIAAD/KeywordExtractor-Datasets/blob/master/datasets/SemEval2017.zip
Dataset Structure
Data Fields
- id: unique identifier of the document.
- document: Whitespace separated list of words in the document.
- doc_bio_tags: BIO tags for each word in the document. B stands for the beginning of a keyphrase and I stands for inside the keyphrase. O stands for outside the keyphrase and represents the word that isn't a part of the keyphrase at all.
- extractive_keyphrases: List of all the present keyphrases.
- abstractive_keyphrase: List of all the absent keyphrases.
Data Splits
Split | #datapoints |
---|---|
Train | 350 |
Test | 100 |
Validation | 50 |
Usage
Full Dataset
from datasets import load_dataset
# get entire dataset
dataset = load_dataset("midas/semeval2017", "raw")
# sample from the train split
print("Sample from train dataset split")
test_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Document BIO Tags: ", test_sample["doc_bio_tags"])
print("Extractive/present Keyphrases: ", test_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", test_sample["abstractive_keyphrases"])
print("\n-----------\n")
# sample from the validation split
print("Sample from validation dataset split")
validation_sample = dataset["validation"][0]
print("Fields in the sample: ", [key for key in validation_sample.keys()])
print("Tokenized Document: ", validation_sample["document"])
print("Document BIO Tags: ", validation_sample["doc_bio_tags"])
print("Extractive/present Keyphrases: ", validation_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", validation_sample["abstractive_keyphrases"])
print("\n-----------\n")
# sample from the test split
print("Sample from test dataset split")
test_sample = dataset["test"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Document BIO Tags: ", test_sample["doc_bio_tags"])
print("Extractive/present Keyphrases: ", test_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", test_sample["abstractive_keyphrases"])
print("\n-----------\n")
Output
Sample from train dataset split
Fields in the sample: ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document: ['It', 'is', 'well', 'known', 'that', 'one', 'of', 'the', 'long', 'standing', 'problems', 'in', 'physics', 'is', 'understanding', 'the', 'confinement', 'physics', 'from', 'first', 'principles.', 'Hence', 'the', 'challenge', 'is', 'to', 'develop', 'analytical', 'approaches', 'which', 'provide', 'valuable', 'insight', 'and', 'theoretical', 'guidance.', 'According', 'to', 'this', 'viewpoint,', 'an', 'effective', 'theory', 'in', 'which', 'confining', 'potentials', 'are', 'obtained', 'as', 'a', 'consequence', 'of', 'spontaneous', 'symmetry', 'breaking', 'of', 'scale', 'invariance', 'has', 'been', 'developed', '[1].', 'In', 'particular,', 'it', 'was', 'shown', 'that', 'a', 'such', 'theory', 'relies', 'on', 'a', 'scale-invariant', 'Lagrangian', 'of', 'the', 'type', '[2]', '(1)L=14w2−12w−FμνaFaμν,', 'where', 'Fμνa=∂μAνa−∂νAμa+gfabcAμbAνc,', 'and', 'w', 'is', 'not', 'a', 'fundamental', 'field', 'but', 'rather', 'is', 'a', 'function', 'of', '4-index', 'field', 'strength,', 'that', 'is,', '(2)w=εμναβ∂μAναβ.', 'The', 'Aναβ', 'equation', 'of', 'motion', 'leads', 'to', '(3)εμναβ∂βw−−FγδaFaγδ=0,', 'which', 'is', 'then', 'integrated', 'to', '(4)w=−FμνaFaμν+M.', 'It', 'is', 'easy', 'to', 'verify', 'that', 'the', 'Aaμ', 'equation', 'of', 'motion', 'leads', 'us', 'to', '(5)∇μFaμν+MFaμν−FαβbFbαβ=0.', 'It', 'is', 'worth', 'stressing', 'at', 'this', 'stage', 'that', 'the', 'above', 'equation', 'can', 'be', 'obtained', 'from', 'the', 'effective', 'Lagrangian', '(6)Leff=−14FμνaFaμν+M2−FμνaFaμν.', 'Spherically', 'symmetric', 'solutions', 'of', 'Eq.', '(5)', 'display,', 'even', 'in', 'the', 'Abelian', 'case,', 'a', 'Coulomb', 'piece', 'and', 'a', 'confining', 'part.', 'Also,', 'the', 'quantum', 'theory', 'calculation', 'of', 'the', 'static', 'energy', 'between', 'two', 'charges', 'displays', 'the', 'same', 'behavior', '[1].', 'It', 'is', 'well', 'known', 'that', 'the', 'square', 'root', 'part', 'describes', 'string', 'like', 'solutions', '[3,4].']
Document BIO Tags: ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'B', 'I', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O']
Extractive/present Keyphrases: ['aaμ equation of motion', 'aναβ equation of motion leads', 'confining part', 'coulomb piece', 'develop analytical approaches', 'quantum theory calculation of the static energy between two charges', 'spherically symmetric solutions', 'spontaneous symmetry breaking of scale invariance', 'string like solutions', 'the effective lagrangian', 'understanding the confinement physics from first principles']
Abstractive/absent Keyphrases: ['(2)w=εμναβ∂μaναβ', 'function of 4-index field strength', 'integrated to (4)w=−fμνafaμν+m', 'leff=−14fμνafaμν+m2−fμνafaμν', 'scale-invariant lagrangian', 'εμναβ∂βw−−fγδafaγδ=0']
-----------
Sample from validation dataset split
Fields in the sample: ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document: ['In', 'the', 'current', 'CLSVOF', 'method,', 'the', 'normal', 'vector', 'is', 'calculated', 'directly', 'by', 'discretising', 'the', 'LS', 'gradient', 'using', 'a', 'finite', 'difference', 'scheme.', 'By', 'appropriately', 'choosing', 'one', 'of', 'three', 'finite', 'difference', 'schemes', '(central,', 'forward,', 'or', 'backward', 'differencing),', 'it', 'has', 'been', 'demonstrated', 'that', 'thin', 'liquid', 'ligaments', 'can', 'be', 'well', 'resolved', 'see', 'Xiao', '(2012).', 'Although', 'a', 'high', 'order', 'discretisation', 'scheme', '(e.g.', '5th', 'order', 'WENO)', 'has', 'been', 'found', 'necessary', 'for', 'LS', 'evolution', 'in', 'pure', 'LS', 'methods', 'to', 'reduce', 'mass', 'error,', 'low', 'order', 'LS', 'discretisation', 'schemes', '(2nd', 'order', 'is', 'used', 'here)', 'can', 'produce', 'accurate', 'results', 'when', 'the', 'LS', 'equation', 'is', 'solved', 'and', 'constrained', 'as', 'indicated', 'above', 'in', 'a', 'CLSVOF', 'method', '(see', 'Xiao,', '2012),', 'since', 'the', 'VOF', 'method', 'maintains', '2nd', 'order', 'accuracy.', 'This', 'is', 'a', 'further', 'reason', 'to', 'adopt', 'the', 'CLSVOF', 'method,', 'which', 'has', 'been', 'used', 'for', 'all', 'the', 'following', 'simulations', 'of', 'liquid', 'jet', 'primary', 'breakup.']
Document BIO Tags: ['O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'B', 'O', 'O', 'B', 'I', 'I', 'B', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O']
Extractive/present Keyphrases: ['5th order weno', 'clsvof method', 'finite difference scheme', 'finite difference schemes', 'high order discretisation scheme', 'liquid', 'low order ls discretisation schemes', 'ls', 'reduce mass error', 'vof method']
Abstractive/absent Keyphrases: ['central, forward, or backward differencing', 'ls methods', 'simulations of liquid jet primary breakup', 'thin liquid ligaments']
-----------
Sample from test dataset split
Fields in the sample: ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document: ['Traditionally,', 'archaeologists', 'have', 'recorded', 'sites', 'and', 'artefacts', 'via', 'a', 'combination', 'of', 'ordinary', 'still', 'photographs,', '2D', 'line', 'drawings', 'and', 'occasional', 'cross-sections.', 'Given', 'these', 'constraints,', 'the', 'attractions', 'of', '3D', 'models', 'have', 'been', 'obvious', 'for', 'some', 'time,', 'with', 'digital', 'photogrammetry', 'and', 'laser', 'scanners', 'offering', 'two', 'well-known', 'methods', 'for', 'data', 'capture', 'at', 'close', 'range', '(e.g.', 'Bates', 'et', 'al.,', '2010;', 'Hess', 'and', 'Robson,', '2010).', 'The', 'highest', 'specification', 'laser', 'scanners', 'still', 'boast', 'better', 'positional', 'accuracy', 'and', 'greater', 'true', 'colour', 'fidelity', 'than', 'SfM–MVS', 'methods', '(James', 'and', 'Robson,', '2012),', 'but', 'the', 'latter', 'produce', 'very', 'good', 'quality', 'models', 'nonetheless', 'and', 'have', 'many', 'unique', 'selling', 'points.', 'Unlike', 'traditional', 'digital', 'photogrammetry,', 'little', 'or', 'no', 'prior', 'control', 'of', 'camera', 'position', 'is', 'necessary,', 'and', 'unlike', 'laser', 'scanning,', 'no', 'major', 'equipment', 'costs', 'or', 'setup', 'are', 'involved.', 'However,', 'the', 'key', 'attraction', 'of', 'SfM–MVS', 'is', 'that', 'the', 'required', 'input', 'can', 'be', 'taken', 'by', 'anyone', 'with', 'a', 'digital', 'camera', 'and', 'modest', 'prior', 'training', 'about', 'the', 'required', 'number', 'and', 'overlap', 'of', 'photographs.', 'A', 'whole', 'series', 'of', 'traditional', 'bottlenecks', 'are', 'thereby', 'removed', 'from', 'the', 'recording', 'process', 'and', 'large', 'numbers', 'of', 'archaeological', 'landscapes,', 'sites', 'or', 'artefacts', 'can', 'now', 'be', 'captured', 'rapidly,', 'in', 'the', 'field,', 'in', 'the', 'laboratory', 'or', 'in', 'the', 'museum.', 'Fig.', '2a–c', 'shows', 'examples', 'of', 'terracotta', 'warrior', 'models', 'for', 'which', 'the', 'level', 'of', 'surface', 'detail', 'is', 'considerable.']
Document BIO Tags: ['O', 'O', 'O', 'O', 'B', 'O', 'B', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'B', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O']
Extractive/present Keyphrases: ['2d line drawings', '3d models', 'archaeological landscapes', 'artefacts', 'control of camera position', 'data capture at close range', 'digital camera', 'digital photogrammetry', 'laser scanners', 'laser scanning', 'ordinary still photographs', 'prior training about the required number and overlap of photographs', 'recording process', 'sfm–mvs', 'sites', 'terracotta warrior models']
Abstractive/absent Keyphrases: ['occasional cross-sections', 'recorded sites and artefacts', 'sfm–mvs methods', 'traditional digital photogrammetry']
-----------
Keyphrase Extraction
from datasets import load_dataset
# get the dataset only for keyphrase extraction
dataset = load_dataset("midas/semeval2017", "extraction")
print("Samples for Keyphrase Extraction")
# sample from the train split
print("Sample from train data split")
test_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Document BIO Tags: ", test_sample["doc_bio_tags"])
print("\n-----------\n")
# sample from the test split
print("Sample from test data split")
test_sample = dataset["test"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Document BIO Tags: ", test_sample["doc_bio_tags"])
print("\n-----------\n")
Keyphrase Generation
# get the dataset only for keyphrase generation
dataset = load_dataset("midas/semeval2017", "generation")
print("Samples for Keyphrase Generation")
# sample from the train split
print("Sample from train data split")
test_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Extractive/present Keyphrases: ", test_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", test_sample["abstractive_keyphrases"])
print("\n-----------\n")
# sample from the test split
print("Sample from test data split")
test_sample = dataset["test"][0]
print("Fields in the sample: ", [key for key in test_sample.keys()])
print("Tokenized Document: ", test_sample["document"])
print("Extractive/present Keyphrases: ", test_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", test_sample["abstractive_keyphrases"])
print("\n-----------\n")
Citation Information
@article{DBLP:journals/corr/AugensteinDRVM17,
author = {Isabelle Augenstein and
Mrinal Das and
Sebastian Riedel and
Lakshmi Vikraman and
Andrew McCallum},
title = {SemEval 2017 Task 10: ScienceIE - Extracting Keyphrases and Relations
from Scientific Publications},
journal = {CoRR},
volume = {abs/1704.02853},
year = {2017},
url = {http://arxiv.org/abs/1704.02853},
eprinttype = {arXiv},
eprint = {1704.02853},
timestamp = {Mon, 13 Aug 2018 16:46:36 +0200},
biburl = {https://dblp.org/rec/journals/corr/AugensteinDRVM17.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
Contributions
Thanks to @debanjanbhucs, @dibyaaaaax and @ad6398 for adding this dataset