Institute of Information Systems Hof University

Institute of Information Systems Hof University

Institute of Information Systems Hof University The institute is a competence centre for the application of information systems in companies. It is the bridge between international research and development and actual application in companies.

Organisation Founded July 2010 Research institute at Hof University 27 staff member (incl. 7 PhD students) 10 20 students

Research Foci Efficient text/data mining Opinion mining Managing heterogeneous data Information systems for business processes System integration Smart grids

Structured data defined structure precise easy to analyse easy to link

Structured data defined structure precise easy to analyse easy to link Unstructured data lacking of defined structures imprecise difficult to analyse difficult to link

Main Application Domains Business information systems GreenTech (e.g. Smart Grids) Civil security/safety

Institute of Information Systems Hof University Dr. Claus Atzenbeck Managing Director Institute of Information Systems Hof University Alfons-Goppel-Platz 1 95028 Hof GERMANY claus.atzenbeck@iisys.de www.iisys.de

Institute of Information Systems Hof University Expertise in greater details by examples

Text/data mining; managing heterogeneous data The Institute of Information Systems provides a high level of expertise in the area management of very large amount of data integrating conventionally structured data with unstructured textual and spatial data. One research focus is intelligent search strategies supported by new types of access structures. On top of these large volumes of heterogeneous data efficient data and text mining methods are investigated (including the detection of topics, named entities and relations). Applications of these methods include: efficient and precise access to the complete data of an organisation, the analysis and combination of heterogeneous data and the deduction of the current state of an organisation as well the as predictions about its future. An example of an ongoing research activity applying these methods is the development of a spatial Internet search engine which was originally developed for security applications. In this context the search engine supports full text search criteria with geographical and temporal criteria to retrieve news related to a certain incident. Future applications of this search engine will support geographical information retrieval in logistic and marketing applications. An ongoing project with a commercial partner implements an information system for managing data on water quality acquired from a large number of measuring points. Beside conventional measuring data also the geographical position of a data sample and explanatory texts have to be stored and analysed. This data will be the basis not only for water quality management but also for complex analytical processes like the tracing of pollutions and the identification of potential origins of a pollution using also text mining methods. Within this domain of expertise the institute is open for new types of applications. This may include but is not limited to business applications like the analysis and optimisation of business processes or the predictions of future market parameter using also unstructured texts from internal and external sources (e.g. the Internet). At present also technical applications are analysed like the prediction of the consumption of electrical energy with a high temporal resolution using not only historical data but also data from public sources.

Aspect-based Opinion Mining; Sentiment Analysis Opinion Mining targets the automatic detection and analysis of opinions included in natural language texts, such as opinions mentioned in blogs, tweets, social media or public news web pages. In particular we do research and development in Aspect based Opinion Mining/Sentiment Analysis for German and English languages, Development of lexical resources (opinion lexicons) and Development of social media monitoring tools Our research may be used in various application domains, such as business (e.g. customer comments on products), politics (e.g. political opinions), finance, medicine, security and many more. For example, our already carried out projects include Politics domain: Politwi (www.politwi.de): In this project we analyze tweets in order to detect controversy discussion and new upcoming themes and topics. Health/medicine domain: Migraine-Radar (www.migraene-radar.de): In this project we harvest tweets containing information about actual migraine attacks and connect them to several other informations to detect correlations, e.g. to weather data. This could be used, for example, to answer the question whether migraine may be caused by certain weather conditions.

Organization Modeling; Business Document Analysis Our research in organizational modelling aims to facilitate the representation of such structures for ICT systems. Many current approaches require a lot of maintenance effort for role definitions as they are unable to efficiently represent primary and secondary organizational structures. The focus of our research in this area is to develop a generic metamodel that is suitable to efficiently represent the multidimensional relations between actors and functional units. This metamodel forms the basis for role management and access administration and allows for their description on an organizational level instead of performing a total enumeration of role members. Applications are: The definition of actors in business processes/workflow management systems Failover strategies when human resources are unavailable Access control on resources In the area of business document analysis, we focus on the automated association of incoming, unstructured documents to business process participants. The aim of this research is to minimize the waiting time for processes to start by improving the lookup of responsible participants. In order to achieve this, statistical text classification methods are enriched with semantic technologies.

Smart Grids The transition from the use of the diminishing fossil or nuclear fuels towards renewable energies does not only imply a change in the generation of electrical power itself, but rather imposes various complex challenges to existing and future energy grids. Among the most prominent challenges are: The volatile nature of, for example, solar or wind energy. Their availability cannot be influenced, leading to substantial fluctuations in the power supplied into the grid. We investigate approaches such as accurate prediction algorithms (based, e.g., on generic algorithms, data mining methodologies, etc.), optimisations of the operating reserve (spinning and non-spinning or replacement reserve), effects of small- and mid-scale energy storage, and remotely controlled consumers and other forms of intelligent load management. The decentralisation of the power generation infrastructure. Solar panels and wind turbines require a substantially lower up-front investment compared to conventional power generation. This leads to a massively increasing complexity of the energy market. Also, leave nodes in the future energy grid may alternate frequently between power supply and power consumption, requiring a bidirectional flow of energy within the network. We seek to develop communication and information flow standards that enable a global load management due to smartly set energy stock exchange incentives that aide in harmonising production and consumption. The conventional methodologies for controlling phase and synchronicity and stabilising network frequency through large rotating masses (spinning reserve) is currently based on large-scale conventional power plants. Due to the decentralisation and small-scale generation, synchronicity becomes a major issue. We research alternative approaches to synchronicity maintenance in the absence of a stabilising spinning reserve based on digital command control. The research group Smart Grids at iisys investigates and develops information technology methodologies and algorithms for future energy grids that carry substantial fractions of renewable energy. This entails the collection and administration of large amounts of production and consumption data, the development of accurate forecasting, algorithms for the planning of the employment of the power generation resources as well as methods for real-time command/control infrastructure. Our focus lies in the development of highly reliable software to guarantee a high level of availability and security for future grids. In the area of smart grids iisys is coordinating the European research network "Reliable Provision of Renewable Energy (RE NET)" that may be used for acquiring additional experts in that field.

2 Project Ideas

Institute of Information Systems Hof University I. Precision Farming

I. Precision Farming Multiple data sources (sensors, Internet) Data fusion Recommendations Predictions

Institute of Information Systems Hof University II. Retroconversion++

Newspaper Archives

Search some keywords

Per son

Org anis atio n

Or n o i t a s i n ga

y t i C

Top ic

Nguyen Dinh Loc [person] Human Rights Watch Viet Tan [organisation] [organisation] HRW opposition suppressed [abbreviation] [topic]

Person A Region C Person B Organisation D Nguyen Dinh Loc [person] Human Rights Watch Viet Tan [organisation] [organisation] opposition suppressed [topic] HRW [abbreviation]

Be a Historian Who has met person A? Who has worked on topic B? Who visited region C? Which topics have been relevant for region D during the time span E? Which topics did person F address? Which persons have been in media very often during the time span G? Which regions have been in the media at what time and on which topics?

End Users Academics (historians, linguists, ) Archives Libraries Retroconversion companies Publishing houses SME Partners wanted