The economic downturn has been forcing many companies to use predictive analysis for spotting emerging product and technology trends and also future customer needs. Since every company is unique, without the assistance of some methodologies and tools, decision makers encounter great difficulties in conducting predictive analysis, especially in the deliberation and prioritization of new prediction criteria derived from the publicly available unstructured information. This paper proposes a unique methodology which attempts to integrate the personalization and visualization of new prediction criteria. The challenging iterative tasks are achieved through a rule-based inconsistency detection triad-based comparison algorithm, supported by sophisticated visual displays of the relative importance among the prediction criteria. It is hoped that the proposed methodology will intuitively support the decision makers in exploring and deliberating new criteria for making better predictions.