MyMessage: case-based reasoning and multicriteria decision making techniques for intelligent context-aware message filtering

摘要

Message filtering is one of the techniques that can help achieve scalability of wireless communication and, eventually, mobile commerce. The core of message filtering is filtering decision-making: to make decisions as to whether a specific message should be filtered or not. The filtering decision is multicriteria decision making in nature, such as analytic hierarchy process (AHP) which ranks a set of alternatives and the choice of a preferred one. Under context-aware situations, however, we are concerned that the number of contextual states is too big, resulting in an AHP model that is too big to compute efficiently. Moreover, in such a situation, user preferences on filtering are dynamic or temporal, according to the change of contexts. Hence, an idea of context-aware message filtering is proposed in this paper. However, to handle a large set of AHP models, each of which fits a specific contextual state, is still difficult. Hence, in this paper, we propose a methodology for intelligent context-aware message filtering by applying a case-based reasoning that learns to adapt to the inexact matching and AHP. Moreover, a default AHP model is generated from the user's personal preference ontology. The default AHP model is refined according to the user's feedback to revise the case base.