Statements: Universal Information Extraction from Tables with Large Language Models for ESG KPIs
Paper
• 2406.19102 • Published
• 1
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 2 new columns ({'input', 'output'}) and 7 missing columns ({'current_hash', 'unstructured_data', 'augmentation_status', 'cell_annotations', 'structured_data', 'ud_type', 'source'}).
This happened while the json dataset builder was generating data using
hf://datasets/docling-project/SemTabNet/se_indirect_1d/train.jsonl (at revision fc268093b21a69f4a04cb45bee67b012a56de66a)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1831, in _prepare_split_single
writer.write_table(table)
File "/usr/local/lib/python3.12/site-packages/datasets/arrow_writer.py", line 714, in write_table
pa_table = table_cast(pa_table, self._schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
input: string
output: string
to
{'structured_data': List({'property': List(Value('string')), 'property_value': List(Value('string')), 'unit': List(Value('string')), 'subject': List(Value('string')), 'subject_value': List(Value('string')), 'predicate_hash': List(Value('string'))}), 'source': {'data_hash': Value('string'), 'document': {'languages': List(Value('string')), 'advanced': {'symbol': Value('string'), 'ticker': Value('string'), 'website': List(Value('string')), 'year': Value('int64'), 'exchange': Value('string'), 'industry': Value('string'), 'location': Value('string'), 'employees': Value('string'), 'sector': Value('string')}, 'subjects': List(Value('string')), 'publication_date': Value('string'), 'affiliations': List({'name': Value('string'), 'id': Value('string'), 'source': Value('string')}), 'collection': {'name': Value('string'), 'alias': List(Value('string')), 'type': Value('string'), 'version': Value('string')}, 'abstract': List(Value('string')), 'url_refs': List(Value('string')), 'title': Value('string'), 'type': Value('string'), 'logs': List({'date': Value('string'), 'agent': Value('string'), 'comment': Value('string'), 'type': Value('string'), 'task': Value('string')}), 'document_hash': Value('string'), 'index_tables': Value('int64')}, 'collection': Value('string'), 'sub_collection': List(Value('string')), 'entities': List(Value('string')), 'ratio_numbers': Value('float64')}, 'cell_annotations': List(List(Value('string'))), 'unstructured_data': List(List(Value('string'))), 'ud_type': Value('string'), 'augmentation_status': Value('string'), 'current_hash': Value('string')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1450, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 993, in stream_convert_to_parquet
builder._prepare_split(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1702, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1833, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 2 new columns ({'input', 'output'}) and 7 missing columns ({'current_hash', 'unstructured_data', 'augmentation_status', 'cell_annotations', 'structured_data', 'ud_type', 'source'}).
This happened while the json dataset builder was generating data using
hf://datasets/docling-project/SemTabNet/se_indirect_1d/train.jsonl (at revision fc268093b21a69f4a04cb45bee67b012a56de66a)
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
structured_data list | source dict | cell_annotations list | unstructured_data list | ud_type string | augmentation_status string | current_hash string |
|---|---|---|---|---|---|---|
[
{
"property": [
"Social: Our People & Our Communities : Corporate Charitable Contributions (excluding employee matching) ",
"time"
],
"property_value": [
"$2,616,880 ",
"FY21 Data "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"sub... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NYSE: PANW",
"ticker": "PANW",
"website": [
"https://www.paloaltonetworks.com/"
],
"year": 2022,
"exchange": "NYSE",
"industry": "Networking & Communication Devices",
"location": "Santa Clara, California",
"employees": "1001-5000 Employees",
"sector": "Technology"
},
"subjects": [
"Technology"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Palo Alto Networks Inc",
"id": "palo-alto",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Palo Alto Networks, Inc. provides enterprise security platform to enterprises, service providers, and government entities worldwide. Its platform includes Next-Generation Firewall that delivers application, user, and content visibility and control, as well as protection against network-based cyber threats; and Threat Intelligence Cloud that offers central intelligence capabilities, as wellas automated delivery of preventative measures against cyber attacks."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NYSE_PANW_2022.pdf"
],
"title": "2022 Environmental, Social and Governance Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T07:02:01.636047+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:26:37.184+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "1edb91af84c6c4c8fce5d2c69f201b2304d01a7a99b6f12b374b48f5af9d0091",
"index_tables": 5
},
"data_hash": "4c71232d4655229f0dcdf987b15c61b7",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"training hours per employee",
"number of other fatalities"
],
"ratio_numbers": 1.1428571428571428
} | [
[
"header_1",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value"
],
[
"property",
"property_value",
"propert... | [
[
"Social: Our People & Our Communities ",
"FY21 Data ",
"FY22 Data "
],
[
"Corporate Charitable Contributions (excluding employee matching) ",
"$2,616,880 ",
"$3,187,997 "
],
[
"Percentage of Employees Participating in Employee Network Groups ",
"25.0% ",
"27.6% "
],
... | table | augmented | fa5d002e9f5147a76ff522c22c6095ae |
[
{
"property": [
"Number of WeSustain teams (global) "
],
"property_value": [
"Not reported "
],
"unit": [
""
],
"subject": [
""
],
"subject_value": [
""
],
"predicate_hash": [
"af4726127c37d038e8ef828731095cf7"
]
},
{
"prope... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NYSE: JNJ",
"ticker": "JNJ",
"website": [
"http://www.jnj.com/"
],
"year": 2021,
"exchange": "NYSE",
"industry": "Drug Manufacturers - Major",
"location": "New Brunswick, New Jersey",
"employees": "10,000+ Employees",
"sector": "Healthcare"
},
"subjects": [
"Healthcare"
],
"publication_date": "2021-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Johnson & Johnson",
"id": "johnson-johnson",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Johnson & Johnson engages in the research and development, manufacture, and sale of various products in the health care field worldwide. The company operates in three segments: Consumer, Pharmaceutical, and Medical Devices and Diagnostics."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NYSE_JNJ_2021.pdf"
],
"title": "2021 Health for Humanity Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T16:28:53.993418+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:27:49.976+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "05b3f5f023c0c8d3a3c8a8d04b55056b2c7d8465a50c88691123dc660f1d1d4e",
"index_tables": 28
},
"data_hash": "2ac81a9fc3c7eb58d66661475fe4a425",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"number of other fatalities"
],
"ratio_numbers": 0.6730769230769231
} | [
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"p... | [
[
"Number of WeSustain teams (global) ",
"Not reported ",
"78 ",
"Not reported "
],
[
"Number of actions taken in 2021 via the HealthyPlanet platform ",
"Not reported ",
"24,609 ",
"Not reported "
],
[
"Number of countries with WeSustain teams ",
"Not reported ",
... | table | augmented | 24acc81f66f6e0503545ca51ba4d676a |
[
{
"property": [
"GHG EMISSIONS SOURCE : Scope 2 Emissions (Market-Based) ",
"time"
],
"property_value": [
"1,544 ",
"2021 GHG EMISSIONS (MTCO2 e ) "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NASDAQ: KFRC",
"ticker": "KFRC",
"website": [
"http://www.kforce.com"
],
"year": 2022,
"exchange": "NASDAQ",
"industry": "Staffing & Outsourcing Services",
"location": "Tampa, Florida",
"employees": "1001-5000 Employees",
"sector": "Services"
},
"subjects": [
"Services"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Kforce Inc.",
"id": "kforce-inc",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Kforce Inc. provides professional and technical specialty staffing services and solutions to the federal government, state and local governments, local and regional companies, and small to mid-sized companies in the United States."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NASDAQ_KFRC_2022.pdf"
],
"title": "2022 Sustainability Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T06:22:53.424664+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:26:04.540+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "9ffc7a0f0930fd60e16b97e905b9851b7c4e8be2bbd4cf932cf037373d19afe9",
"index_tables": 0
},
"data_hash": "3710aa4499519c61b9dad91bbd5b4838",
"collection": null,
"sub_collection": null,
"entities": null,
"ratio_numbers": null
} | [
[
"header_1",
"time_value",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_valu... | [
[
"GHG EMISSIONS SOURCE ",
"2021 GHG EMISSIONS (MTCO2 e ) ",
"2019 GHG EMISSIONS (MTCO2 e ) ",
"2020 GHG EMISSIONS (MTCO2 e ) "
],
[
"Scope 2 Emissions (Market-Based) ",
"1,544 ",
"2,236 ",
"1,815 "
],
[
"Total Scope 1 & Scope 2 Emissions (Market-Based) ",
... | table | augmented | d1037cb8375977fb182a3d2168209c79 |
[
{
"property": [
"INDICATORS : Number and percentage of workers covered by an occupational health and safety management system ",
"time"
],
"property_value": [
"100% ",
"2020 "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "ASX: YAL",
"ticker": "YAL",
"website": [
"https://www.yancoal.com.au/"
],
"year": 2022,
"exchange": "ASX",
"industry": "Thermal Coal",
"location": "Sydney, Australia",
"employees": "1001-5000 Employees",
"sector": "Energy"
},
"subjects": [
"Energy"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Yancoal Australia Ltd",
"id": "yancoal-australia-ltd",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Yancoal Australia Ltd engages in the exploration, development, production, and marketing of metallurgical and thermal coal in Japan, Singapore, China, South Korea, Taiwan, Thailand, Australia, and internationally. It owns 95% interests in the Moolarben coal mine located in the Western Coalfields of New South Wales; 100% interests in the Stratford Duralie mines located within the New SouthWales Gloucester Basin; 100% interests in the Yarrabee mine located to the northeast of Blackwater in Central Queensland's Bowen Basin; and 80% interests in the Mount Thorley mine and 84.5% interests in the Warkworth mine in the Hunter Valley region of New South Wales. It also manages six other projects in New South Wales, Queensland, and Western Australia. The company was founded in 2004 and is based in Sydney, Australia. Yancoal Australia Ltd is a subsidiary of Yanzhou Coal Mining Company Limited."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/ASX_YAL_2022.pdf"
],
"title": "2022 ESG Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T12:33:27.584590+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:31:29.419+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "401b916dbabd69c1590296d7d5a1860d087bd4932c9e58ce3bf150e145e0d782",
"index_tables": 13
},
"data_hash": "10ba73d1770dff1693cec42947520406",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"number of other fatalities"
],
"ratio_numbers": 0.8095238095238095
} | [
[
"header_1",
"time_value",
"time_value",
"time_value",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"p... | [
[
"INDICATORS ",
"2020 ",
"2019 ",
"2022 ",
"2018 ",
"2021 "
],
[
"Number and percentage of workers covered by an occupational health and safety management system ",
"100% ",
"100% ",
"100% ",
"100% ",
"100% "
],
[
"Number of hours worked ",
" 8,200,7... | table | augmented | b59417f6d3c97da6cd60e12acccb1ea3 |
[
{
"property": [
"Metrics : Total number of people reached (cumulative, FY20 to current reporting FY) ",
"Notes ",
"time"
],
"property_value": [
"46,588,226 ",
"Our “1 Billion Lives” goal tracks the number of people who are reached through our health and education initiativ... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NYSE: DELL",
"ticker": "DELL",
"website": [
"https://www.dell.com/en-us"
],
"year": 2022,
"exchange": "NYSE",
"industry": "Information Technology Services",
"location": "Round Rock, Texas",
"employees": "10,000+ Employees",
"sector": "Technology"
},
"subjects": [
"Technology"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Dell Technologies Inc.",
"id": "dell-technologies-inc",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Dell Technologies is a unique family of businesses that provides the essential infrastructure for organizations to build their digital future, transform IT and protect their most important asset, information. The company services customers of all sizes across 180 countries â ranging from 98 percent of the Fortune 500 to individual consumers â with the industry's most comprehensive and innovativeportfolio from the edge to the core to the cloud."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NYSE_DELL_2022.pdf"
],
"title": "2022 Environmental, Social and Governance Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-16T12:45:05.289782+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:36:38.651+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "357afbc6a87e1b9e8c3da44277f40b59e2adbcadc463f504983959b7356cd11f",
"index_tables": 22
},
"data_hash": "3ee9c70cf28cc51e7925863db8283fff",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"number of other fatalities"
],
"ratio_numbers": 1.2285714285714286
} | [
[
"header_1",
"time_value",
"time_value",
"time_value",
"key"
],
[
"property",
"property_value",
"property_value",
"property_value",
"key_value"
],
[
"property",
"property_value",
"property_value",
"property_value",
"empty"
],
[
"property",... | [
[
"Metrics ",
"FY20 ",
"FY21 ",
"FY22 ",
"Notes "
],
[
"Total number of people reached (cumulative, FY20 to current reporting FY) ",
"46,588,226 ",
"93,565,402 ",
"159,742,242 ",
"Our “1 Billion Lives” goal tracks the number of people who are reached through our health a... | table | augmented | e31cfcb5d2e59396f41424d704564886 |
[
{
"property": [
"Emissions Intensity : Emissions Intensity (MT CO2- e / USD Revenue) ",
"time"
],
"property_value": [
"0.0003 ",
"2021 "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
""
],
... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NYSE: GRA",
"ticker": "GRA",
"website": [
"http://www.grace.com/"
],
"year": 2022,
"exchange": "NYSE",
"industry": "Specialty Chemicals",
"location": "Columbia, Maryland",
"employees": "1001-5000 Employees",
"sector": "Basic Materials"
},
"subjects": [
"Basic Materials"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "W. R. Grace and Company",
"id": "wr-grace-company",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Built on talent, technology, and trust, Grace is a leading global supplier of catalysts and engineered materials. The company's two industry-leading business segments \"Catalysts Technologies and Materials Technologies\" provide innovative products, technologies, and services that enhance the products and processes of there customer partners around the world."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NYSE_GRA_2022.pdf"
],
"title": "2022 GRI Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T21:40:08.082809+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:28:23.536+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "8cd65dce75249647d91213f414f3ca0c0b735bde6b01d38e40e0efcfb099df48",
"index_tables": 11
},
"data_hash": "972d35dd7fa4867ac72549425ad56799",
"collection": "esg",
"sub_collection": [
"environment"
],
"entities": [
"scope 1 ghg emissions"
],
"ratio_numbers": 1.5
} | [
[
"header_1",
"time_value",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_valu... | [
[
"Emissions Intensity ",
"2021 ",
"2020 ",
"2019 "
],
[
"Emissions Intensity (MT CO2- e / USD Revenue) ",
"0.0003 ",
"0.0003 ",
"0.0003 "
],
[
"Total Revenue (including ART Joint Venture, USD) ",
"2,512,600,000 ",
"2,211,900,000 ",
"2,468,500,000 "... | table | augmented | 3c16678348a858366676387937f237d8 |
[
{
"property": [
"Scope 1 (direct emissions), including: ",
"time"
],
"property_value": [
"682,645 ",
"2021 "
],
"unit": [
"t of CO$_{2}$e ",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
""
],
"predicate_ha... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "ASX: POL",
"ticker": "POL",
"website": [
"https://polymetals.com/"
],
"year": 2022,
"exchange": "ASX",
"industry": "Gold",
"location": "Sydney, Australia",
"employees": "11-50 Employees",
"sector": "Basic Materials"
},
"subjects": [
"Basic Materials"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Polymetals Resources Ltd",
"id": "polymetal-resources-ltd",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Polymetals Resources Ltd (Polymetals) aims to become a gold production company, initially focussing on its two 100% owned exploration licences within Guineaâs Siguiri Basin, totalling 112km2. The Siguiri Basin occupies the north-eastern corner of Guinea and hosts several large active gold mining operations and is notable for its widespread gold anomalism."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/ASX_POL_2022.pdf"
],
"title": "2022 Sustainability Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T03:01:16.364532+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:33:11.551+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "14b8115e1829aa9bb7ad83f79c50b71d296fe32274b1abc6235663719730c4f4",
"index_tables": 203
},
"data_hash": "d9fafd8e7ca1610af2e3f68b448797c2",
"collection": "esg",
"sub_collection": [
"environment"
],
"entities": [
"scope 1 ghg emissions",
"scope 3 ghg emissions"
],
"ratio_numbers": 1.2901234567901234
} | [
[
"empty",
"unit_property",
"time_value",
"time_value",
"time_value",
"time_value"
],
[
"property",
"unit_value",
"property_value",
"property_value",
"property_value",
"property_value"
],
[
"property",
"unit_value",
"property_value",
"property_... | [
[
"",
"Units ",
"2021 ",
"2020 ",
"2022 ",
"2019 "
],
[
"Scope 1 (direct emissions), including: ",
"t of CO$_{2}$e ",
"682,645 ",
"612,669 ",
"751,486 ",
"613,717 "
],
[
"Combustion of fuels in stationary sources, including: ",
"t of CO$_{2}$e ",
"... | table | augmented | 0dc9b1b27eaa9047af395184a2285ead |
[
{
"property": [
"Training and Development : Number of external people who received Tarkett Academy training ",
"GRI ",
"time"
],
"property_value": [
"8,148 ",
"203-2 ",
"2021 "
],
"unit": [
"",
"",
""
],
"subject": [
"",
"... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "OTC: TKFTF",
"ticker": "TKFTF",
"website": [
"http://www.tarkett-group.com/"
],
"year": 2022,
"exchange": "OTC",
"industry": "Building Materials Wholesale",
"location": "Paris, France",
"employees": "10,000+ Employees",
"sector": "Basic Materials"
},
"subjects": [
"Basic Materials"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Tarkett",
"id": "tarkett",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"With a history of more than 140 years, Tarkett is a worldwide leader in innovative and sustainable flooring and sports surface solutions, generating net sales of ⬠3.4 billion in 2022. The Group employs 12,000 employees and has 25 R&D centers, 8 recycling centers and 34 production sites. Tarkett creates and manufactures solutions for hospitals, schools, housing, hotels, offices, storesand sports fields, serving customers in over 100 countries. To build âThe Way to Better Floors,â the Group is committed to circular economy and sustainability, in line with its Tarkett HumanâConscious Design® approach."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/OTC_TKFTF_2022.pdf"
],
"title": "2022 Corporate Social & Environmental Responsibility Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-16T11:46:48.737217+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:36:01.839+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "94d2c28cfd1f37e1f2e80850364f349bb846aef0816e34557e2a258c36182c7c",
"index_tables": 21
},
"data_hash": "955da44628c12a6b974db08c2725c26d",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"training hours per employee"
],
"ratio_numbers": 0.5576923076923077
} | [
[
"key",
"header_1",
"time_value",
"time_value",
"time_value",
"time_value",
"time_value",
"time_value"
],
[
"empty",
"header_1",
"header_1",
"header_1",
"header_1",
"header_1",
"header_1",
"header_1"
],
[
"key_value",
"property",
"... | [
[
"GRI ",
"Indicator ",
"2021 ",
"2019 ",
"2020 ",
"Variation 2022 vs. base year ",
"2022 ",
"Variation 2022 vs. 2021 "
],
[
"",
"Training and Development ",
"Training and Development ",
"Training and Development ",
"Training and Development ",
"Trainin... | table | augmented | b26ba0ecf0846ec853d03b848ce465e3 |
[
{
"property": [
"Average training hours per year/employee ",
"category"
],
"property_value": [
"23.5 ",
"Male "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
""
],
"predicate_hash": [
... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NYSE: CSTM",
"ticker": "CSTM",
"website": [
"https://www.constellium.com/"
],
"year": 2022,
"exchange": "NYSE",
"industry": "Aluminum",
"location": "Paris, France",
"employees": "10,000+ Employees",
"sector": "Basic Materials"
},
"subjects": [
"Basic Materials"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Constellium SE",
"id": "constellium-se",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Constellium is a leading full-service supplier of rolled and extruded aluminium solutions for the global automotive market. They help engineer lighter, stronger and safer vehicles to deliver higher performance, lower emissions, better fuel economy and greater range. With a product offer that includes solutions for closures, structural components, Crash Management Systems, battery enclosures, decorative parts and heat exchangers, Constellium aluminium enables advanced mobility. One in four vehicles produced in Europe and North America in 2019 features aluminium from Constellium."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NYSE_CSTM_2022.pdf"
],
"title": "2022 Sustainability Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T02:03:55.046604+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:24:07.669+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "3edb14748d577ed6cce573b95c9748f6254e75d6ab7147415e83baf188ff01d0",
"index_tables": 11
},
"data_hash": "2b49845b9d0ab5c4fdc28f9a5e2becef",
"collection": "esg",
"sub_collection": [
"social"
],
"entities": [
"training hours per employee"
],
"ratio_numbers": 0
} | [
[
"empty",
"key_value",
"key_value",
"key_value",
"key_value",
"key_value"
],
[
"property",
"property_value",
"property_value",
"property_value",
"property_value",
"property_value"
]
] | [
[
"",
"Male ",
"All employees ",
"Managers 1 ",
"Operators 2 ",
"Female "
],
[
"Average training hours per year/employee ",
"23.5 ",
"23.8 ",
"23.1 ",
"24.0 ",
"26.1 "
]
] | table | augmented | 586ddd875ba42ee5164ac53e9afb6018 |
[
{
"property": [
"Greenhouse Gas (“GHG”) Emissions Summary (tCO2e) : GHG Emissions by Scope : Scope 3 ",
"time"
],
"property_value": [
"438,731 ",
"2020 "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "NASDAQ: ETSY",
"ticker": "ETSY",
"website": [
"https://www.etsy.com/"
],
"year": 2022,
"exchange": "NASDAQ",
"industry": "Specialty Retail, Other",
"location": "rooklyn, New York",
"employees": "501-1000 Employees",
"sector": "Services"
},
"subjects": [
"Services"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Etsy Inc",
"id": "etsy-inc",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Etsy, Inc. operates online and offline marketplaces to buy and sell handmade items, vintage goods, and craft supplies. Its platform connects sellers and buyers to sell or buy products for art, home and living, mobile accessories, jewelry, wedding, and others. The company was founded in 2005 and is headquartered in Brooklyn, New York. It has additional offices in Berlin, Germany; Dublin, Ireland; Hudson, New York; London, United Kingdom; Melbourne, Australia; Paris, France; San Francisco, California; and Toronto, Canada."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/NASDAQ_ETSY_2022.pdf"
],
"title": "2022 Responsibility Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T09:59:42.340016+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:26:37.185+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "260b43d5948d8a19876338bf55047ebb0d93680f064c2aacfcde1fa723fa5ede",
"index_tables": 13
},
"data_hash": "bd325d80596c5afb09c0eeaef9f01ef3",
"collection": "esg",
"sub_collection": [
"environment"
],
"entities": [
"scope 3 ghg emissions",
"scope 1 ghg emissions"
],
"ratio_numbers": 0.7916666666666666
} | [
[
"header_1",
"rubbish",
"rubbish",
"rubbish"
],
[
"header_2",
"time_value",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value",
"property_value"
],
[
"property",
"empty",
"property_value",
"empty"
],
[
"... | [
[
"Greenhouse Gas (“GHG”) Emissions Summary (tCO2e) ",
"Etsy.com + Reverb ",
"Etsy.com, Reverb, Depop ",
"Etsy.com + Reverb "
],
[
"GHG Emissions by Scope ",
"2020 ",
"2022 ",
"2021 "
],
[
"Scope 3 ",
"438,731 ",
"531,638 ",
"548,900 "
],
[
"Scope ... | table | augmented | 55f0aacb0f325abcdf59091c016d8070 |
[
{
"property": [
"GREENHOUSE GAS EMISSIONS : CO$_{2}$ emissions : Gross global Scope 2 emissions location-based ",
"SASB code ",
"time"
],
"property_value": [
"20 ",
"Additional ",
"Data 2022 "
],
"unit": [
"Metric tons CO$_{2}$-e ",
"",
... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "OTC: PRSEY",
"ticker": "PRSEY",
"website": [
"https://www.prosafe.com/"
],
"year": 2022,
"exchange": "OTC",
"industry": "Oil & Gas Equipment & Services",
"location": "Stavanger, Norway",
"employees": "51-200 Employees",
"sector": "Energy"
},
"subjects": [
"Energy"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Prosafe Offshore Pte Ltd",
"id": "prosafe-offshore-pte-ltd",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Prosafe SE, together with its subsidiaries, owns and operates semi-submersible accommodation vessels in Europe, South America, and internationally. The company is also involved in the maintenance and modification of installations on fields for the production, hook-up and commissioning of new fields, tie-backs to existing infrastructure, and decommissioning activities. It owns and operatesa fleet of seven semi-submersible accommodation vessels and one tender support vessel. The company primarily serves oil and gas industries. Prosafe SE was founded in 1972 and is headquarter in Stavanger, Norway."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/OTC_PRSEY_2022.pdf"
],
"title": "2022 Environmental, Social & Governance Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T16:19:14.231380+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:33:11.551+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "63c14db47a7db04bdaba43c09357c4645749541c5f43d68db386e779a8b8605f",
"index_tables": 15
},
"data_hash": "b0e0847e99e2f1adb7133ca1cc00fc33",
"collection": "esg",
"sub_collection": [
"environment"
],
"entities": [
"scope 1 ghg emissions",
"scope 3 ghg emissions"
],
"ratio_numbers": 0.7052631578947368
} | [
[
"header_1",
"unit_property",
"time_value",
"time_value",
"key"
],
[
"header_1",
"header_1",
"header_1",
"header_1",
"header_1"
],
[
"header_2",
"header_2",
"header_2",
"header_2",
"header_2"
],
[
"property",
"unit_value",
"propert... | [
[
"Accounting metric ",
"Unit of measure ",
"Data 2022 ",
"Data 2021 ",
"SASB code "
],
[
"GREENHOUSE GAS EMISSIONS ",
"GREENHOUSE GAS EMISSIONS ",
"GREENHOUSE GAS EMISSIONS ",
"GREENHOUSE GAS EMISSIONS ",
"GREENHOUSE GAS EMISSIONS "
],
[
"CO$_{2}$ emissi... | table | augmented | 23ab652a70c23882b9777061884f2114 |
[
{
"property": [
"INDICATORS : Total Emissions tCO2-e ",
"time"
],
"property_value": [
"1,983,298 ",
"2019 "
],
"unit": [
"",
""
],
"subject": [
"",
""
],
"subject_value": [
"",
""
],
"predicate_hash": [
"7e... | {
"document": {
"languages": [
"en"
],
"advanced": {
"symbol": "ASX: YAL",
"ticker": "YAL",
"website": [
"https://www.yancoal.com.au/"
],
"year": 2022,
"exchange": "ASX",
"industry": "Thermal Coal",
"location": "Sydney, Australia",
"employees": "1001-5000 Employees",
"sector": "Energy"
},
"subjects": [
"Energy"
],
"publication_date": "2022-01-01T12:00:00.000+00:00",
"affiliations": [
{
"name": "Yancoal Australia Ltd",
"id": "yancoal-australia-ltd",
"source": "ir-solutions"
}
],
"collection": {
"name": "ESG Reports",
"alias": [
"esg-report"
],
"type": "Document",
"version": "1.3.0"
},
"abstract": [
"Yancoal Australia Ltd engages in the exploration, development, production, and marketing of metallurgical and thermal coal in Japan, Singapore, China, South Korea, Taiwan, Thailand, Australia, and internationally. It owns 95% interests in the Moolarben coal mine located in the Western Coalfields of New South Wales; 100% interests in the Stratford Duralie mines located within the New SouthWales Gloucester Basin; 100% interests in the Yarrabee mine located to the northeast of Blackwater in Central Queensland's Bowen Basin; and 80% interests in the Mount Thorley mine and 84.5% interests in the Warkworth mine in the Hunter Valley region of New South Wales. It also manages six other projects in New South Wales, Queensland, and Western Australia. The company was founded in 2004 and is based in Sydney, Australia. Yancoal Australia Ltd is a subsidiary of Yanzhou Coal Mining Company Limited."
],
"url_refs": [
"https://www.responsibilityreports.com/HostedData/ResponsibilityReports/PDF/ASX_YAL_2022.pdf"
],
"title": "2022 ESG Report",
"type": "ESG report",
"logs": [
{
"date": "2023-08-17T12:33:27.584590+00:00",
"agent": "CCS",
"comment": "CCS v9.1.1 parsing of documents",
"type": "parsing",
"task": null
},
{
"date": "2023-08-21T11:31:29.419+00:00",
"agent": "CXS",
"comment": "enrich esg-report-metadata-20230628 to esg-report-20230815",
"type": "text enrichment",
"task": "enrich esg-report-metadata-20230628 to esg-report-20230815"
}
],
"document_hash": "401b916dbabd69c1590296d7d5a1860d087bd4932c9e58ce3bf150e145e0d782",
"index_tables": 3
},
"data_hash": "bae0ebdc666cea92089e352bb56711b8",
"collection": "esg",
"sub_collection": [
"environment"
],
"entities": [
"scope 1 ghg emissions"
],
"ratio_numbers": 1.8055555555555556
} | [
[
"header_1",
"time_value",
"time_value",
"time_value",
"time_value",
"time_value"
],
[
"property",
"property_value",
"property_value",
"property_value",
"property_value",
"property_value"
],
[
"property",
"property_value",
"property_value",
"p... | [
[
"INDICATORS ",
"2019 ",
"2018 ",
"2022 ",
"2021 ",
"2020 "
],
[
"Total Emissions tCO2-e ",
"1,983,298 ",
"2,114,527 ",
"2,367,913 ",
"2,213,876 ",
"2,042,183 "
],
[
"Total ROM production ",
"51,574,833 ",
"49,455,204 ",
"46,507,466 ",
"... | table | augmented | 50bce7f8897edc85980d3bac0772d7bd |
End of preview.
YAML Metadata Warning: The task_categories "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other
Dataset Card for SemTabNet
This dataset accompanies the following paper:
Title: Statements: Universal Information Extraction from Tables with Large Language Models for ESG KPIs
Authors: Lokesh Mishra, Sohayl Dhibi, Yusik Kim, Cesar Berrospi Ramis, Shubham Gupta, Michele Dolfi, Peter Staar
Venue: Accepted at the NLP4Climate workshop in the 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024)
In this paper, we propose STATEMENTS as a new knowledge model for storing quantiative information in a domain agnotic, uniform structure. The task of converting a raw input (table or text) to Statements is called Statement Extraction (SE). The statement extraction task falls under the category of universal information extraction.
- Code Repository: SemTabNet repository
- Arxiv Paper: Statements: Universal Information Extraction from Tables with Large Language Models for ESG KPIs
- Point of Contact: IBM Research DeepSearch Team
Data Splits
There are three tasks supported by this dataset. The data for each three task is split in training, validation, and testing set. Additionally, we also provide the original annotations of the raw tables which are used to construct all other data.
| Task | Train | Test | Valid |
|---|---|---|---|
| SE Direct | 103455 | 11682 | 5445 |
| SE Indirect 1D | 72580 | 8489 | 3821 |
| SE Indirect 2D | 93153 | 22839 | 4903 |
Languages
The text in the dataset is in English.
Source and Annotations
The source of this dataset and the annotation strategy is described in the paper.
Citation Information
Arxiv: https://arxiv.org/abs/2406.19102
- Downloads last month
- 116