Artikel in Tagungsband INPROC-2016-06

Bibliograph.
Daten
Kassner, Laura; Mitschang, Bernhard: Exploring Text Classification for Messy Data: An Industry Use Case for Domain-Specific Analytics.
In: Advances in Database Technology - EDBT 2016, 19th International Conference on Extending Database Technology, Bordeaux, France, March 15-16, Proceedings.
Universität Stuttgart, Fakultät Informatik, Elektrotechnik und Informationstechnik.
S. 491-502, englisch.
OpenProceedings.org, 9. März 2016.
ISBN: 978-3-89318-070-7.
Artikel in Tagungsband (Konferenz-Beitrag).
KörperschaftEDBT
CR-Klassif.H.3.1 (Content Analysis and Indexing)
H.3.3 (Information Search and Retrieval)
H.4.2 (Information Systems Applications Types of Systems)
J.1 (Administration Data Processing)
Keywordsrecommendation system; automotive; text analytics; domain-specific language; automatic classification
Kurzfassung

Industrial enterprise data present classification problems which are different from those problems typically discussed in the scientific community -- with larger amounts of classes and with domain-specific, often unstructured data. We address one such problem through an analytics environment which makes use of domain-specific knowledge. Companies are beginning to use analytics on large amounts of text data which they have access to, but in day-to-day business, manual effort is still the dominant method for processing unstructured data. In the face of ever larger amounts of data, faster innovation cycles and higher product customization, human experts need to be supported in their work through data analytics. In cooperation with a large automotive manufacturer, we have developed a use case in the area of quality management for supporting human labor through text analytics: When processing damaged car parts for quality improvement and warranty handling, quality experts have to read text reports and assign error codes to damaged parts. We design and implement a system to recommend likely error codes based on the automatic recognition of error mentions in textual quality reports. In our prototypical implementation, we test several methods for filtering out accurate recommendations for error codes and develop further directions for applying this method to a competitive business intelligence use case.

Volltext und
andere Links
Volltext-PDF
KontaktEmail an laura.kassner@ipvs.uni-stuttgart.de
Abteilung(en)Universität Stuttgart, Institut für Parallele und Verteilte Systeme, Anwendersoftware
Projekt(e)GSaME C2-001 ApPLAUDING
Eingabedatum7. April 2016
   Publ. Abteilung   Publ. Institut   Publ. Informatik