Method for Analyzing User Action Logs (tool LogAnalyzer).

A method for discovering of user characteristics based on analysis of recorded sequences of user actions.

Institution: Slovak University of Technology
Technologies used: Java, Nazou Integration Technology, Ontological interface of Nazou Corporate Memory, Hibernate, MySQL
Inputs: Log of user actions
Outputs: User characteristics in user model
Documentation: HTML, doc, JavaDoc
Distribution packages: zip
Video: demonstration video

Addressed Problems

Adaptive web based system are adaptating to user features. For this, they need to have a user model filled with appropriate characteristics. One possible way of automatic acquisition of user characteristics is to monitor user behavior within the system and create a log of user actions.

However, this log does not provide a direct information about user characteristics, so we have to determine meanings are either implied by or associated with the user behavior (e.g., to identify user goals or estimate the knowledge about a certain concept)

Description

A method uses pre-defined navigational patterns which occur in collected data. These patterns are determined from a model of navigation on a web site. Afterwards, it uses heuristics defined according to the application domain of a system. In our case, application domain is online labor market, where users attempt to find a most suitable offer. Heuristics thus connect the knowledge about user goals (find a job) with a knowledge about user navigation (choose of different facets etc.) to make an estimation of some user characteristics in a model.

The method can determine user characteristics also from user reactions upon displayed content using user feedback. For this, it needs to compare two information objects (job offers in our domain) and get their common and different aspects to find out why user reacted differently upon them. For this purpose, additional comparing methods tools like Method for Ontology Concept and Instance Comparison (tool ConCom) and Method for Ontology Concept and Instance Similarity Measurement (tool OntoSim) were defined.

Log Analyzer

Overview of a process of automatic acquisition of user characteristics

References

  1. Barla, M.: Interception of User's Interests on the Web. In V. Wade, H. Ashman, and B. Smyth, editors, 4th Int. Conf. on Adaptive Hypermedia and Adaptive Web-Based Systems, AH'06, pages 435-439, Dublin, Ireland, 2006, Springer, LNCS 4018
  2. Barla, M., Andrejko, A., Bieliková, M., Tvarožek, M.: User Characteristics Acquisition from Logs with SemanticsIn ISIM ´07 Information Systems and Formal Models: 10th International Conference on Information System Implementation and Modeling. 2nd International Workshop on Formal Models WFM ´07 2007, Hradec nad Moravicí, Czech Republic, pp.103-110, 2007
  3. Barla, M., Bieliková, M.: Estimation of User Characteristics using Rule-based Analysis of User Logs. In: Data Mining for User Modeling Proceedings of Workshop held at the International Conference on User Modeling UM2007, Corfu, Greece, pp.5-14, 2007