Complex adaptive filtering user profile using graphical models

Abstract

This article explores how to develop complex data driven user models that go beyond the bag of words model and topical relevance. We propose to learn from rich user specific information and to satisfy complex user criteria under the graphical modelling framework. We carried out a user study with a web based personal news filtering system, and collected extensive user information, including explicit user feedback, implicit user feedback and some contextual information. Experimental results on the data set collected demonstrate that the graphical modelling approach helps us to better understand the complex domain. The results also show that the complex data driven user modelling approach can improve the adaptive information filtering performance. We also discuss some practical issues while learning complex user models, including how to handle data noise and the missing data problem.

Pattern-based causal relationships discovery from event sequences for modeling behavioral user profile in ubiquitous environments

“…Zhang [68] proposed a model for constructing the user profile using a graphical approach. His model is based on independence relationships and outputs an equivalent Bayesian network.…”

“…His model is based on independence relationships and outputs an equivalent Bayesian network. The work in [68] does not take into account the variability of user behavior, and assumes that a user always exhibits the same behavior in performing activities.…”

Air-combat behavior data mining based on truncation method

“…Gagné et al presented a robust estimation and prediction in multivariate autoregressive models with exogenous variables, and tested several scenarios based on the finite sample properties of the robust prediction intervals in simulation [7]. Zhang proposed an approach to learn from rich user specific information and to satisfy complex user criteria under the graphical modeling framework, and the experimental results that demonstrated this approach was helpful to better understand the complex domain [8]. Gancarski et al proposed two new feature weighting methods based on coevolutive algorithms to process the complex data, and the experimental results showed that the methods were better than the hill-climbing based algorithms [9].…”