The study finds that: (1) The content platform presents two empowerment models based on different recommendation scenarios of recommendation algorithms, content oriented and community oriented, each with distinct value propositions. (2) Recommendation algorithms play a key role in business model innovation, with different algorithm empowerment mechanisms for each model. Specifically, the content oriented model manifests as causal recommendation and associated recommendation, whereas the community oriented model manifests as emotional recommendation and social recommendation. (3) The value created by platforms under different algorithm empowerment models varies. The content oriented model achieves the purpose of user drainage through recommendation algorithm empowerment, accumulating a large user base and generating user acquisition value for the platform. The community oriented model achieves the purpose of user retention through recommendation algorithm empowerment, enhancing user stickiness to the platform and generating user retention value.