Hugging Face's logo

Datasets:
midas
/
inspec

Modalities:
Text
Size:
1K - 10K
ArXiv:
Libraries:
Datasets
Dataset card Viewer Files Community
inspec / README.md
dibyaaaaax's picture
dibyaaaaax
Update README.md
82813e8
preview code
|
raw
history blame
15.9 kB

Dataset Summary

A dataset for benchmarking keyphrase extraction and generation techniques from abstracts of english scientific papers. For more details about the dataset please refer the original paper - https://dl.acm.org/doi/pdf/10.3115/1119355.1119383 Original source of the data -

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 1000
Test 500
Validation 500

Usage

Full Dataset 

from datasets import load_dataset

# get entire dataset
dataset = load_dataset("midas/inspec", "raw")

# sample from the train split
print("Sample from training dataset split")
train_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in train_sample.keys()])
print("Tokenized Document: ", train_sample["document"])
print("Document BIO Tags: ", train_sample["doc_bio_tags"])
print("Extractive/present Keyphrases: ", train_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", train_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 training data split
Fields in the sample:  ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document:  ['A', 'conflict', 'between', 'language', 'and', 'atomistic', 'information', 'Fred', 'Dretske', 'and', 'Jerry', 'Fodor', 'are', 'responsible', 'for', 'popularizing', 'three', 'well-known', 'theses', 'in', 'contemporary', 'philosophy', 'of', 'mind', ':', 'the', 'thesis', 'of', 'Information-Based', 'Semantics', '-LRB-', 'IBS', '-RRB-', ',', 'the', 'thesis', 'of', 'Content', 'Atomism', '-LRB-', 'Atomism', '-RRB-', 'and', 'the', 'thesis', 'of', 'the', 'Language', 'of', 'Thought', '-LRB-', 'LOT', '-RRB-', '.', 'LOT', 'concerns', 'the', 'semantically', 'relevant', 'structure', 'of', 'representations', 'involved', 'in', 'cognitive', 'states', 'such', 'as', 'beliefs', 'and', 'desires', '.', 'It', 'maintains', 'that', 'all', 'such', 'representations', 'must', 'have', 'syntactic', 'structures', 'mirroring', 'the', 'structure', 'of', 'their', 'contents', '.', 'IBS', 'is', 'a', 'thesis', 'about', 'the', 'nature', 'of', 'the', 'relations', 'that', 'connect', 'cognitive', 'representations', 'and', 'their', 'parts', 'to', 'their', 'contents', '-LRB-', 'semantic', 'relations', '-RRB-', '.', 'It', 'holds', 'that', 'these', 'relations', 'supervene', 'solely', 'on', 'relations', 'of', 'the', 'kind', 'that', 'support', 'information', 'content', ',', 'perhaps', 'with', 'some', 'help', 'from', 'logical', 'principles', 'of', 'combination', '.', 'Atomism', 'is', 'a', 'thesis', 'about', 'the', 'nature', 'of', 'the', 'content', 'of', 'simple', 'symbols', '.', 'It', 'holds', 'that', 'each', 'substantive', 'simple', 'symbol', 'possesses', 'its', 'content', 'independently', 'of', 'all', 'other', 'symbols', 'in', 'the', 'representational', 'system', '.', 'I', 'argue', 'that', 'Dretske', "'s", 'and', 'Fodor', "'s", 'theories', 'are', 'false', 'and', 'that', 'their', 'falsehood', 'results', 'from', 'a', 'conflict', 'IBS', 'and', 'Atomism', ',', 'on', 'the', 'one', 'hand', ',', 'and', 'LOT', ',', 'on', 'the', 'other']
Document BIO Tags:  ['O', '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', 'B', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'B', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'B', 'O', 'B', 'O', 'O', 'O', 'O', 'O', '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', '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', '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', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O']
Extractive/present Keyphrases:  ['philosophy of mind', 'content atomism', 'ibs', 'language of thought', 'lot', 'cognitive states', 'beliefs', 'desires']
Abstractive/absent Keyphrases:  ['information-based semantics']

-----------

Sample from validation data split
Fields in the sample:  ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document:  ['Impact', 'of', 'aviation', 'highway-in-the-sky', 'displays', 'on', 'pilot', 'situation', 'awareness', 'Thirty-six', 'pilots', '-LRB-', '31', 'men', ',', '5', 'women', '-RRB-', 'were', 'tested', 'in', 'a', 'flight', 'simulator', 'on', 'their', 'ability', 'to', 'intercept', 'a', 'pathway', 'depicted', 'on', 'a', 'highway-in-the-sky', '-LRB-', 'HITS', '-RRB-', 'display', '.', 'While', 'intercepting', 'and', 'flying', 'the', 'pathway', ',', 'pilots', 'were', 'required', 'to', 'watch', 'for', 'traffic', 'outside', 'the', 'cockpit', '.', 'Additionally', ',', 'pilots', 'were', 'tested', 'on', 'their', 'awareness', 'of', 'speed', ',', 'altitude', ',', 'and', 'heading', 'during', 'the', 'flight', '.', 'Results', 'indicated', 'that', 'the', 'presence', 'of', 'a', 'flight', 'guidance', 'cue', 'significantly', 'improved', 'flight', 'path', 'awareness', 'while', 'intercepting', 'the', 'pathway', ',', 'but', 'significant', 'practice', 'effects', 'suggest', 'that', 'a', 'guidance', 'cue', 'might', 'be', 'unnecessary', 'if', 'pilots', 'are', 'given', 'proper', 'training', '.', 'The', 'amount', 'of', 'time', 'spent', 'looking', 'outside', 'the', 'cockpit', 'while', 'using', 'the', 'HITS', 'display', 'was', 'significantly', 'less', 'than', 'when', 'using', 'conventional', 'aircraft', 'instruments', '.', 'Additionally', ',', 'awareness', 'of', 'flight', 'information', 'present', 'on', 'the', 'HITS', 'display', 'was', 'poor', '.', 'Actual', 'or', 'potential', 'applications', 'of', 'this', 'research', 'include', 'guidance', 'for', 'the', 'development', 'of', 'perspective', 'flight', 'display', 'standards', 'and', 'as', 'a', 'basis', 'for', 'flight', 'training', 'requirements']
Document BIO Tags:  ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', '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', 'O', 'B', 'I', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', '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', 'O', 'B', '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', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O']
Extractive/present Keyphrases:  ['flight simulator', 'pilots', 'cockpit', 'flight guidance', 'situation awareness', 'flight path awareness']
Abstractive/absent Keyphrases:  ['highway-in-the-sky display', 'human factors', 'aircraft display']

-----------

Sample from test data split
Fields in the sample:  ['id', 'document', 'doc_bio_tags', 'extractive_keyphrases', 'abstractive_keyphrases', 'other_metadata']
Tokenized Document:  ['A', 'new', 'graphical', 'user', 'interface', 'for', 'fast', 'construction', 'of', 'computation', 'phantoms', 'and', 'MCNP', 'calculations', ':', 'application', 'to', 'calibration', 'of', 'in', 'vivo', 'measurement', 'systems', 'Reports', 'on', 'a', 'new', 'utility', 'for', 'development', 'of', 'computational', 'phantoms', 'for', 'Monte', 'Carlo', 'calculations', 'and', 'data', 'analysis', 'for', 'in', 'vivo', 'measurements', 'of', 'radionuclides', 'deposited', 'in', 'tissues', '.', 'The', 'individual', 'properties', 'of', 'each', 'worker', 'can', 'be', 'acquired', 'for', 'a', 'rather', 'precise', 'geometric', 'representation', 'of', 'his', '-LRB-', 'her', '-RRB-', 'anatomy', ',', 'which', 'is', 'particularly', 'important', 'for', 'low', 'energy', 'gamma', 'ray', 'emitting', 'sources', 'such', 'as', 'thorium', ',', 'uranium', ',', 'plutonium', 'and', 'other', 'actinides', '.', 'The', 'software', 'enables', 'automatic', 'creation', 'of', 'an', 'MCNP', 'input', 'data', 'file', 'based', 'on', 'scanning', 'data', '.', 'The', 'utility', 'includes', 'segmentation', 'of', 'images', 'obtained', 'with', 'either', 'computed', 'tomography', 'or', 'magnetic', 'resonance', 'imaging', 'by', 'distinguishing', 'tissues', 'according', 'to', 'their', 'signal', '-LRB-', 'brightness', '-RRB-', 'and', 'specification', 'of', 'the', 'source', 'and', 'detector', '.', 'In', 'addition', ',', 'a', 'coupling', 'of', 'individual', 'voxels', 'within', 'the', 'tissue', 'is', 'used', 'to', 'reduce', 'the', 'memory', 'demand', 'and', 'to', 'increase', 'the', 'calculational', 'speed', '.', 'The', 'utility', 'was', 'tested', 'for', 'low', 'energy', 'emitters', 'in', 'plastic', 'and', 'biological', 'tissues', 'as', 'well', 'as', 'for', 'computed', 'tomography', 'and', 'magnetic', 'resonance', 'imaging', 'scanning', 'information']
Document BIO Tags:  ['O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'B', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'B', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'I', 'I', 'I', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'B', 'O', 'B', 'I', 'O', 'O', 'B', 'I', 'I', 'I', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'B', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'B', 'O', 'O', 'B', 'O', 'O', 'O', 'O', 'B', 'I', 'O', 'B', 'I', 'I', 'I', 'I']
Extractive/present Keyphrases:  ['computational phantoms', 'monte carlo calculations', 'in vivo measurements', 'radionuclides', 'tissues', 'worker', 'precise geometric representation', 'mcnp input data file', 'scanning data', 'computed tomography', 'brightness', 'graphical user interface', 'computation phantoms', 'calibration', 'in vivo measurement systems', 'signal', 'detector', 'individual voxels', 'memory demand', 'calculational speed', 'plastic', 'magnetic resonance imaging scanning information', 'anatomy', 'low energy gamma ray emitting sources', 'actinides', 'software', 'automatic creation']
Abstractive/absent Keyphrases:  ['th', 'u', 'pu', 'biological tissues']

-----------

Keyphrase Extraction 

from datasets import load_dataset

# get the dataset only for keyphrase extraction
dataset = load_dataset("midas/inspec", "extraction")

print("Samples for Keyphrase Extraction")

# sample from the train split
print("Sample from training data split")
train_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in train_sample.keys()])
print("Tokenized Document: ", train_sample["document"])
print("Document BIO Tags: ", train_sample["doc_bio_tags"])
print("\n-----------\n")

# sample from the validation split
print("Sample from validation data 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("\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/inspec", "generation")

print("Samples for Keyphrase Generation")

# sample from the train split
print("Sample from training data split")
train_sample = dataset["train"][0]
print("Fields in the sample: ", [key for key in train_sample.keys()])
print("Tokenized Document: ", train_sample["document"])
print("Extractive/present Keyphrases: ", train_sample["extractive_keyphrases"])
print("Abstractive/absent Keyphrases: ", train_sample["abstractive_keyphrases"])
print("\n-----------\n")

# sample from the validation split
print("Sample from validation data 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("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 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 

@inproceedings{hulth2003improved,
  title={Improved automatic keyword extraction given more linguistic knowledge},
  author={Hulth, Anette},
  booktitle={Proceedings of the 2003 conference on Empirical methods in natural language processing},
  pages={216--223},
  year={2003}
}

Contributions

Thanks to @debanjanbhucs, @dibyaaaaax and @ad6398 for adding this dataset