Learning and Deducing Temporal Orders

This paper studies how to determine temporal orders on attribute values in a set of tuples that pertain the same entity, absence complete timestamps. We propose creator-critic framework learn and deduce by combining deep learning rule-based deduction, referred as GATE (Get lATEst). The creator trains ranking model via learning, rank based correlations among attributes. critic then validates learned deduces more ranked pairs chasing data with currency constraints; it also provides augmented training feedback for improve next round. process proceeds until order obtained becomes stable. Using real-life synthetic datasets, we show is able F -measure above 80%, improving 7.8% methods 34.4%.