This repository provides reources developed within the following paper:

S. Zhang and K. Balog. EntiTables: Smart Assistance for Entity-Focused Tables. - SIGIR'17

This study is an effort aimed at reproducing the result presented in the Smart table paper.

This repository is structured as follows:

• Data: The table corpus is WikiTables, which comprises 1.6M tables extracted from Wikipedia. We proproceeed it and make it public downloadable here.
• Population: All the core evaluation of population tasks are provided here.
• Output: The output files can only be requested by email now.

The data we used are public data sets:

• DBpedia 2015-10
• WikiTable from http://websail-fe.cs.northwestern.edu/TabEL/

[NOTE] We are using elastic 2 ( > 2.3), elasticsearch 5 will encounter some minor problems with elastic.py wrapper. To score the column labels, we need to build a table index with multiple fields using elasticsearch. An exmaple indexer is provided for indexing. Index your table corpus data following this example and start your population:)

@inproceedings{Zhang:2017:ESA,
 author = {Zhang, Shuo and Balog, Krisztian},
 title = {EntiTables: Smart Assistance for Entity-Focused Tables},
 booktitle = {Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval},
 series = {SIGIR '17},
 year = {2017},
 isbn = {978-1-4503-5022-8},
 location = {Shinjuku, Tokyo, Japan},
 pages = {255--264},
 numpages = {10},
 url = {http://doi.acm.org/10.1145/3077136.3080796},
 doi = {10.1145/3077136.3080796},
 acmid = {3080796},
 publisher = {ACM},
 address = {New York, NY, USA},
 keywords = {intelligent table assistance, semantic search, table completion},
}

If you have any question, please contact Shuo Zhang at [email protected] or Krisztian Balog at [email protected]