SemanticLog

A tool for effective logging and integration of semantic events from various sources.

Institution: Slovak University of Technology
Technologies used: Java, XML, Web Services, SQL (planned)
Inputs: Event ontology, semantic events corresponding to user actions
Outputs: Log with the supplied semantic events (XML file and/or SQL database).
Documentation: HTML, doc, JavaDoc

Addressed Problems

User modeling with minimal user involvement and automatic acquisition of user characteristics require comprehensive logs of a user's interaction with the system. In principle, various data can be acquired and subsequently logged by sever-side tools, which can further be used to determine user characteristics. Simple logs of e.g. URL addresses however provide only limited information about what had actually happened in a system and thus make it difficult to infer sensible user characteristics.

Furthermore, if several interacting server-side tools cooperate within a system, creating a comprehensive log of user activity becomes more difficult while the use of client-side logging tools makes logging and log integration a serious problem. Additionally log analysis tools depend on individual system tools because the logging of semantics of user actions and their corresponding server events requires a well defined cooperation and support of multiple tools.

Description

The basic principle of the semantic log lies in the use of an event ontology that describes the events that might occur as a result of user actions performed while using a system. Various server-side or client-side tools that acquire data can access a common web service interface and store data about events and the associated semantics in a convenient way in one place. Furthermore, events from various sources are integrated into a single stream of semantic events (per user), which then provides simple and effective access to the stored data for different data mining tools (e.g. for user characteristics analysis). The use of a common event ontology decreases the dependency of user modeling tools on individual system tools since they no more need to support the specifics of each tool. Some examples of events that can be logged include the selection of facet restrictions in Factic (Faceted Semantic Browser) or the displaying of details for a particular publication.

Thus, a semantic logging tool allows client tools to perform the following operations:

  • Access a common web service interface
  • Logging of events and their associated semantics
  • Integration of various data sources on a per user basis
  • Consistent access for data mining tools to source data from different client tools
Overview of integration of SemanticLog with other tools.

Overview of integration of SemanticLog with other tools.

References

  1. Tvarožek, M., Barla, M., Bieliková, M. (2007). Personalized Presentation in Web-Based Information Systems. In J. Van Leeuwen, G. F. Italiano, W. van der Hoek, H. Sack, C. Meinel, F. Plášil (Ed.), SOFSEM 2007: Proceedings of the 33rd Conference on Current Trends in Theory and Practice of Computer Science. LNCS 4362, pp. 796-807. Harrachov, Czech Republic: Springer-Verlag, Berlin Heidelberg.
  2. Andrejko, A., Barla, M., Bieliková, M., Tvarožek, M. (2006). Softvérové nástroje pre získavanie charakteristík používateľa. In P. Vojtáš & T. Skopal (Ed.), Proceedings of DATAKON '06, (pp. 139-148). Brno, Czech Republic (in Slovak).