Behavioural similarity analysis for supporting the recommendation in PAbMM
Mario José Diván, María Laura Sánchez Reynoso · 2017
The transactions through Internet continuously happen and represents a propitious field for the data stream processing. Moreover, the decision making process associated with the online processing, requires support or knowledge from previous experiences. With the aim of to be more accuracy in the managing and retrieving of the organizational knowledge (or previous experiences), we introduce a model for the behavioural similarity analysis from similar structurally entities based on a measurement and evaluation framework. This analysis is added to the structural similarity analysis in the processing architecture based on measurement metadata for better supporting of the recommendation based on the organizational memory. The behavioural analysis allows limiting the search space of available recommendations in the organizational memory and to be more accuracy for recommending courses of action. It is a key asset in the processing of the measurement streams, on the one hand, before situations in which there not exists specific previous experiences related to an entity under monitoring; on the other hand, considering that the organizational memory is hosted on a big data ecosystem.