Using an asset price bubble model in tweet analytics

Predictive methods in the context of big data needs to adopt new approaches. Justification can be derived from the volume of public data entering through microblogs in short periods of time. Based on such data predictions and other inferences are made in many areas such as politics, entertainment, and emergencies. In this paper we present B-function threshold model (BTM), a new approach for predictive analytics. BTM consists of a function/series (B-function) and a threshold. This approach is to predefine an appropriate model and test whether the data being considered meets threshold condition. Inferences are drawn based on the comparison. In this paper we adopt an asset bubble model that can capture mean reversion, stochastic, and speculative regimes as B-function. Empirical analysis is shown by applying the proposed method to Twitter data.