Hugging Face's logo

Datasets:
sap-ai-research
/
SALT

Modalities:
Text
Formats:
parquet
Size:
1M - 10M
Libraries:
Datasets
Dask
License:
Dataset card Viewer Files Community
1
Dataset Viewer
Auto-converted to Parquet
Full Screen
Search is not available for this dataset
ADDRESSID
stringlengths
10
10
⌀
ADDRESSREPRESENTATIONCODE
stringclasses
1 value
COUNTRY
stringclasses
240 values
REGION
stringclasses
597 values
__index_level_0__
int64
0
2.32M
0050000002
DE
1,575,149
0050000005
DE
1,575,082
0050000007
DE
371,313
0050000013
DE
06
1,575,192
0050000014
DE
326,792
0050000015
DE
1,574,464
0050000017
DE
06
1,574,795
0050000018
DE
06
1,575,081
0050000019
DE
06
1,574,778
0050000052
DE
06
1,575,297
0050000056
DE
370,547
0050000057
DE
370,548
0050000067
DE
06
1,575,142
0050000068
DE
748,157
0050000069
DE
06
1,575,226
0050000070
DE
748,091
0050000071
DE
749,754
0050000075
DE
747,875
0050000079
DE
06
1,575,227
0050000080
DE
1,574,573
0050000083
DE
747,704
0050000108
DE
747,608
0050000128
DE
229,722
0050000168
DE
747,143
0050000188
DE
747,673
0050000228
DE
747,333
0050000308
DE
747,196
0050000348
DE
747,292
0050000388
DE
1,575,147
0050000408
DE
747,675
0050000409
DE
747,170
0050000448
DE
577,994
0050000648
DE
747,311
0050000769
NL
1,252,098
0050000782
NL
1,051,132
0050000788
DE
747,688
0050000868
DE
747,455
0050000948
DE
747,224
0050000969
BE
1,783,330
0050000988
DE
747,187
0050001008
DE
747,186
0050001010
BE
1,667,863
0050001069
NL
1,631,821
0050001072
NL
1,632,642
0050001148
DE
747,422
0050001408
DE
749,755
0050002109
DE
747,160
0050002110
DE
747,674
0050002149
DE
747,856
0050002250
DE
747,312
0050002390
DE
747,291
0050002430
DE
747,434
0050002470
DE
06
747,690
0050002489
GB
1,652,955
0050002510
DE
747,416
0050002570
DE
06
747,691
0050002635
GB
1,638,921
0050002784
BE
481,097
0050002917
GB
1,638,916
0050002943
GB
1,652,971
0050003010
NL
419,430
0050003011
NL
419,468
0050003014
NL
990,828
0050003052
NL
991,448
0050003053
NL
991,386
0050003054
NL
991,311
0050003055
NL
991,340
0050003057
NL
991,536
0050003058
NL
991,229
0050003093
BE
481,073
0050003094
NL
419,420
0050003095
GR
493,112
0050003096
AE
350,112
0050003097
FR
427,200
0050003098
PT
504,934
0050003099
NL
419,421
0050003100
DE
370,438
0050003101
IT
443,862
0050003102
ES
541,168
0050003103
NO
484,230
0050003104
SE
549,319
0050003105
BE
481,074
0050003106
BE
481,075
0050003107
BE
481,076
0050003108
BE
481,077
0050003112
NL
991,018
0050003113
BE
1,002,165
0050003115
NL
991,562
0050003116
NL
991,302
0050003117
NL
990,823
0050003118
BE
1,002,258
0050003119
NL
990,772
0050003122
NL
991,400
0050003123
NL
990,739
0050003132
HU
551,237
0050003193
NL
991,410
0050003194
NL
991,070
0050003212
DE
747,172
0050003293
BE
1,002,875
0050003353
BE
1,002,877
End of preview. Expand in Dataset Viewer.

SALT: Sales Autocompletion Linked Business Tables Dataset

License

Dataset for our paper SALT: Sales Autocompletion Linked Business Tables Dataset presented at NeurIPS'24 Table Representation Workshop.

News

  • 12/15/2024: Preliminatry dataset now also available on Hugging Face. Train/test split will be provided in the next days
  • 12/13/2024: Provided data
  • 10/29/2024: Preliminary repository created

Abstract

Foundation models, particularly those that incorporate Transformer architectures, have demonstrated exceptional performance in domains such as natural language processing and image processing. Adapting these models to structured data, like tables, however, introduces significant challenges. These difficulties are even more pronounced when addressing multi-table data linked via foreign key, which is prevalent in the enterprise realm and crucial for empowering business use cases. Despite its substantial impact, research focusing on such linked business tables within enterprise settings remains a significantly important yet underexplored domain. To address this, we introduce a curated dataset sourced from an Enterprise Resource Planning (ERP) system, featuring extensive linked tables. This dataset is specifically designed to support research endeavors in table representation learning. By providing access to authentic enterprise data, our goal is to potentially enhance the effectiveness and applicability of models for real-world business contexts.

Information

Table Schema of SALT Dataset Table Schema of SALT Dataset

Screenshot of a Salesorder Input Mask Example Input Mask of a Salesorder App using SAP S/4HANA

Usage

Example of loading the tables with pandas. Unless already installed, install it with: 

pip install pandas

import pandas as pd

# load the table data from the parquet files
salesdocuments = pd.read_parquet("data/I_SalesDocument.parquet")
salesdocument_items = pd.read_parquet("data/I_SalesDocumentItem.parquet")
customers = pd.read_parquet("data/I_Customer.parquet")
addresses = pd.read_parquet("data/I_AddrOrgNamePostalAddress.parquet")

# show the first elements
salesdocuments.head()

Authors:

Citations

If you use this code in your research or want to refer to our work, please cite:

@inproceedings{
klein2024salt,
title={{SALT}: Sales Autocompletion Linked Business Tables Dataset},
author={Tassilo Klein and Clemens Biehl and Margarida Costa and Andre Sres and Jonas Kolk and Johannes Hoffart},
booktitle={NeurIPS 2024 Third Table Representation Learning Workshop},
year={2024},
url={https://openreview.net/forum?id=UZbELpkWIr}
}

Roadmap

  • Integration into RelBench, Feb'25
  • Release dataset
Downloads last month
10
Size of downloaded dataset files:
68.3 MB
Size of the auto-converted Parquet files:
68.3 MB
Number of rows:
4,749,350