Bilingual information processing has been being used widely in many tasks such as Natural Language Understanding (NLU), Cross-Domain, Machine Translation (MT) and Bilingual Cognition. There still remain some problematic phenomena for understanding bilingual information processing. In this work, we aim to fully understand both the principles and methods of bilingual information processing by taking advantage of appropriate comparisons. The major idea of our proposed approach is both human and the machine can cooperate or interact with each other according to their characteristics. Concretely, it is composed of the following three steps: First, find issues from the Chinese-English bilingual translation at phrase sentence level alignment. The challenge in Chinese-English translation lies in an inconsistency of the meanings between the Chinese characters and the English words. Then, analyze the problems from the view of interdisciplinary, cross-domain and cross-industry. Finally, address leveraging human-machine to divide the task and coordinate with each other. As a result, not only we discover both parallel dual formalized approach of big and small strings, but also find and validate both the superiority and followed scientific principles of generalized bilingual information processing method.