Concept Drift Awareness in Twitter Streams

摘要: Learning in non-stationary environments is not an easy task and requires a distinctive approach. The learning model must only have the ability to continuously learn, but also acquired new concepts forget old ones. Additionally, given significant importance that social networks gained as information networks, there ever-growing interest extraction of complex used for trend detection, promoting services or market sensing. This dynamic nature tends limit performance traditional static models strategies be put forward. In this paper we present strategy learn with drift occurrence Twitter. We propose three different models: time-window model, ensemble-based incremental model. Since little known about types can occur Twitter, simulate by artificially time stamping real Twitter messages order evaluate validate our strategy. Results are so far encouraging regarding presence drift, along classifying streams.