Logo des Repositoriums
 

Towards Learned Metadata Extraction for Data Lakes

dc.contributor.authorLangenecker, Sven
dc.contributor.authorSturm, Christoph
dc.contributor.authorSchalles, Christian
dc.contributor.authorBinnig, Carsten
dc.contributor.editorKai-Uwe Sattler
dc.contributor.editorMelanie Herschel
dc.contributor.editorWolfgang Lehner
dc.date.accessioned2021-03-16T07:57:10Z
dc.date.available2021-03-16T07:57:10Z
dc.date.issued2021
dc.description.abstractAn important task for enabling the efficient exploration of available data in a data lake is to annotate semantic type information to the available data sources. In order to reduce the manual overhead of annotation, learned approaches for automatic metadata extraction on structured data sources have been proposed recently. While initial results of these learned approaches seem promising, it is still not clear how well these approaches can generalize to new unseen data in real-world data lakes. In this paper, we aim to tackle this question and as a first contribution show the result of a study when applying Sato -a recent approach based on deep learning -to a real-world data set. In our study we show that Sato is only able to extract semantic data types for about 10% of the columns of the real-world data set. These results show the general limitation of deep learning approaches which often provide near-perfect performance on available training and testing data but fail in real settings since training data and real data often strongly vary. Hence, as a second contribution we propose a new direction of using weak supervision and present results of an initial prototype we built to generate labeled training data with low manual efforts to improve the performance of learned semantic type extraction approaches on new unseen data sets.en
dc.identifier.doi10.18420/btw2021-17
dc.identifier.isbn978-3-88579-705-0
dc.identifier.pissn1617-5468
dc.identifier.urihttps://dl.gi.de/handle/20.500.12116/35800
dc.language.isoen
dc.publisherGesellschaft für Informatik, Bonn
dc.relation.ispartofBTW 2021
dc.relation.ispartofseriesLecture Notes in Informatics (LNI) - Proceedings, Volume P-311
dc.subjectdata lakes
dc.subjectdataset discovery and search
dc.subjectsemantic type detection
dc.titleTowards Learned Metadata Extraction for Data Lakesen
gi.citation.endPage336
gi.citation.startPage325
gi.conference.date13.-17. September 2021
gi.conference.locationDresden
gi.conference.sessiontitleData Integration, Semantic Data Management, Streaming

Dateien

Originalbündel
1 - 1 von 1
Vorschaubild nicht verfügbar
Name:
A3-23.pdf
Größe:
949.06 KB
Format:
Adobe Portable Document Format