University of Hamburg - to research, to teach, to educate and form, to homepage
Language Technology Group (LT)
Department of InformaticsLanguage Technology Group (LT)

Photo: alt left

Language Technology

ABSA-DB: Aspect-based Sentiment Analysis for DB Products and Services

Description

Subject of this project is the examination of the feasibility of a technology prototype for aspect-based sentiment analysis.  DB plans to evolve into a usable in practice software in the event of a positive test and based on the contractual prototypical development.
The main objective of the individual project is the creation of a methodology for the automatic aggregation of opinions from German free text fields of customer questionnaires and social media regarding individual products and services of DB. This will enable DB to analyse  the reception of advertising campaigns and increase customer satisfaction through faster response ability.
The projects enables LT at TU Darmstadt to methodologically advance the detection of aspect-based sentiment in German texts, provides access to realistic, German sentiment texts, as well as an evaluation under real conditions. Research questions exist in the adaptation of sentiment analysis for German texts to DB-themes and in the general methodology for the establishment of an aspect of inventory, identification of issues and the targeted selection of manually annotated examples.

Goals

The prototype developed in this project shall show a satisfactory performance on aspect-based sentiment detection for German texts for DB aspects. This will be achieved by syntactic processing of German, semantic methods of language technologies and machine learning for structured prediction. A crucial element is the creation of an annotated aspect-level dataset of significant size, which a) characterizes identified in cooperation with the DB aspect classes , b) the opinion indicating elements maps the aspects and c) represents topics of the source texts representatively. The prototype will be available as open source software under a permissive license.

Project Data

Funding Body: Deutsche Bahn, within the Innovation alliance DB & TU Darmstadt
Project volume: 60K Euro
Project Duration: Feb 2016 - Dec 2016

Project Partners

1. Deutsche Bahn Mobility Logistics AG, Frankfurt, Germany
2. Technische Universität Darmstadt, FG Language Technology, Darmstadt, Germany

People

Publications

  • Kumar, A., Kohail, S., Ekbal, A., Biemann C. (2015): IIT-TUDA: System for Sentiment Analysis in Indian Languages using Lexical Acquisition. In: Third International Conference on Mining Intelligence and Knowledge Exploration (MIKE 2015). Hyderabad, India (pdf)
  • Wojatzki M., Ruppert E., Zesch T., Biemann C. (Editors) (2017): GermEval 2017: Shared Task on Aspect-based Sentiment in Social Media Customer Feedback. Proceedings of the GSCL GermEval Shared Task on Aspect-based Sentiment in Social Media Customer Feedback, September 12, 2017, Berlin, Germany. (pdf)
  • Wojatzki M., Ruppert E., Holschneider S., Zesch T., Biemann C. (2017): GermEval 2017: Shared Task on Aspect-based Sentiment in Social Media Customer Feedback. In Proceedings of the GSCL GermEval Shared Task on Aspect-based Sentiment in Social Media Customer Feedback, pages 1--12, Berlin, Germany. (pdf)
  • Ruppert E., Kumar A., Biemann C. (2017): LT-ABSA: An extensible open-source system for document-level and aspect-based sentiment analysis. In Proceedings of the GSCL GermEval Shared Task on Aspect-based Sentiment in Social Media Customer Feedback, pages 55–60, Berlin, Germany. (pdf)