Factorization forecasting approach for user modeling


  • Nguyễn Thái Nghe Can Tho University
  • Lars Schmidt-Thieme Information Systems and Machine Learning Lab Marienburger Platz 22 University of Hildesheim 31141 Hildesheim




User modeling, matrix factorization, factorization forecasting, sequential effect, recommender systems, intelligent tutoring systems


User modeling is a task which customizes and adapts the systems to meet users' specific needs. The user modeling is widely used in many areas.  For example, in e-commerce, it is used for modeling consumers' preferences (behaviors) then predicting their future preferences to recommend suitable products to them. In e-learning (e.g., intelligent tutoring systems - ITS), the user modeling is used to model the learners (students) to track/predict their performance/knowledge.

In this work, an approach which integrates forecasting model into matrix factorization model to take into account sequential/temporal effects in user modeling since users' need/knowledge may change overtime is introduced. The model as well as how to use stochastic gradient descent to learn this model, then resulting with an algorithm are thoroughly presented.

The proposed model is validated using several data sets which are extracted from both e-commerce and e-learning areas. Experimental results on these data sets show that the proposed approach performs nicely. This could be a promising approach for user modeling.


Metrics Loading ...

Author Biographies

Nguyễn Thái Nghe, Can Tho University

Nguyen Thai-Nghe is a lecturer at the Department of In-formation Systems, College of Information and Communication Technology, Can Tho University, Vietnam. He received his BSc. in Informatics in 1999 at Cantho University and obtained the degree of Master of Engineering in Information Management at Asian Institute of Technology, Thailand, in 2006. He received his PhD at the University of Hildesheim, Germany, in 2012. His main interest research is machine learning, especially in clas-sication for imbalanced data, data mining and recommender systems, especially for education. He has published his papers in 2 book chapters (Springer and IGI Global) and several international conferences (e.g., IEEE-FIE, IEEE-ISDA, IEEE-ICALT, IEEE-RIVF, ACM-SoICT,..).

Lars Schmidt-Thieme, Information Systems and Machine Learning Lab Marienburger Platz 22 University of Hildesheim 31141 Hildesheim

Lars Schmidt-Thieme is full professor leading the Information Systems and Machine Learning Lab (ISMLL), University of Hildesheim, Germany. He obtained his Diploma in Mathematics at the University of Heidelberg in 1999 and received his PhD at the University of Karlsruhe, Germany, in 2003. From 2003 to 2006 he was professor at the Institute for Computer Science atthe University of Freiburg, Germany. His main research interest is machine learning, especially classication and regression problems as well as recommender systems. He has published articles in top journals (Springer, ACM, IEEE) and top international conferences (e.g. IEEE ICDM, ACM KDD, UAI). He is member of program committees of international top conferences (e.g. ACM KDD, SIAM SDM, ECML/PKDD) as well asexecutive board member of the German Classication Society.




How to Cite

N. T. Nghe and L. Schmidt-Thieme, “Factorization forecasting approach for user modeling”, JCC, vol. 31, no. 2, pp. 133–148, Jun. 2015.



Computer Science