While deceptive behaviors are a natural part ofhuman life, it is well known that human is generally bad at detecting deception. In this study, we present an automatic deception detection framework by comprehensively integratingprior domain knowledge in deceptive behavior understanding. Specifically, we compute acoustics, textual information, implicatures with non-verbal behaviors, and conversational temporal dynamics for improvingautomatic deception detection in dialogs. The proposed model reaches start-of-the-art performance on the Daily Deceptive Dialogues corpus ofMandarin (DDDM) database, 80.61% unweighted accuracy recall in deception recognition. In the further analyses, we reveal that (i) the deceivers' deception behaviors can be observed fromthe interrogators' behaviors in the conversational temporal dynamics features and (ii) some ofthe acoustic features (e.g. loudness andMFCC) and textual features are significant and effective indicators to detect deception behaviors.