Logo

General Online Research 2011

March 14-16, 2011, Heinrich-Heine University of Düsseldorf

DGOF Logo

Conference Agenda

Overview and details of the sessions of this conference. Please select a date or room to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).

 
Session Overview
Session

BP2: Best Practice Award II

Time: Tuesday, 15/Mar/2011: 4:00pm - 5:00pm
Session Chair: Holger Geißler

Presentations

Insight Mining & Insight News: Using Web 2.0 Analytic Methods for Data Management Systems and News Providing

Stefan Althoff1, Peter Gentsch2

1Lufthansa Technik AG, Germany; 2TextTech GmbH, Germany

Relevance & Research Question: One of the key factors for the analysis of user generated content (UGC) is the usage of full automated textmining routines. Due to the amount of text generated in the Web every day the usage of such techniques is the only chance to handle this valuable information source. But is the usage of such sophisticated methods – including single procedures like tokenization, part-of-speech-tagging, disassembly, term extraction – limited to UGC only?

Methods & Data: Textmining can be used for a data management system based on offline sources as well. Nevertheless it is possible to build up a data pool filled with offline documents and in addition with online data like newsletter, website content and UGC. The big advantage of textmining usage for dynamic data management systems with multiple sources: The storing structure for the documents is no object; and by implementing a graphic interface and displaying (statistical) relations between terms and concepts by a semantic net it is possible to explore the data pool in a knowledge generating matter. Regularly recurring search requests can be implemented in addition, e.g. for a regular news screening.

Results: These are the ideas behind the data management system Insight Mining (roll-out: 2008) and its spin-off Insight News (roll-out: 2011); both tools were jointly developed by TextTech (Leipzig) and Lufthansa Technik. (Hamburg). Both concepts will be presented in this session.

Added Value: Data analyzes and news screening can be done more efficient and more economical compared to other known systems.

Contact : Stefan Althoff can be contacted during the conference via mobile phone ( +49 (0)151 589 162 41 ) for further questions and discussions. Please do not hesitate to contact him


Quick and clever – creating an automated facebook monitoring tool for MAM

Andreas Woppmann1, Karl Ledermüller2

1MAM Babyartikel GmbH, Austria; 2Vienna University of Economics and Business

• Relevance & Research Question: MAM Babyartikel is an international producer of babies products and developed the first pacifier together with designers. As a brand with focus on design and state of the art products, it is important to us that our consumers see MAM as a brand, which is modern and open minded for new communication channels. Doubtlessly social networks play an important role in getting and staying in touch with customers. For this purpose marketing resources are bound in facebook related activities. A continuous qualitative evaluation regarding the efficiency of this marketing channel is quite expensive, especially when creating benchmarks for activities in subsidiaries and competitors. Therefore we developed an R and LaTeX based benchmarking environment for open access to facebook data.

• Methods & Data:

Data were harvested over the facebook API and a set of developed metrics and plots was calculated. Additionally text mining(tm) and common methods in tm (MDS or clustering) helped to give an overview of topics of interest in each of the monitored subsidiary and competitor sites. Statistics were calculated with the help of R. Monthly and quarterly reports were generated with the in Stanford developed document mark up language LaTeX.

• Results: With standardised reports, we created the base for benchmarking of the facebook accounts of our subsidiaries. Additionally we use the insights in the different markets for implicit trend monitoring. Hence, we are able to help subsidiaries not only with information about benchmarks against competitors, but also about which topics are highly discussed by our facebook community worldwide. Quantitative benchmarking against competitors is extremely helpful to find best practice examples in facebook communication. Thanks to it we are able to improve the communication with our community. With the standardisation of the report we achieved a cheap and fast monitoring of our Facebook sides.

• Added Value: The project helped us and our subsidiaries to use facebook based communication more efficiently with a moderate project budget. Additionally we are able to monitor product relevant ongoing discussions which help us to get an insight into topics which are relevant for our community.

Woppmann-Quick and clever – creating an automated facebook monitoring tool-215.pptx

 
Imprint · Contact Address:
Conference: GOR 2011
Conference Software - ConfTool Pro 2.6.17
© 2001 - 2010 by H. Weinreich, Hamburg, Germany