Devices and edges of the Internet of Things (IoT) are working together to broaden the sensing capability of the IoT. The protection of personal sensory information privacy in the smart home has gotten more attention. To enable secure data sharing under the scenario of the smart home, attribute-based encryption (ABE) has been applied widely. As a general rule, the sophisticated calculation of ABE is difficult under resource-constrained devices. The appearance of edge computing alleviates this dilemma, where an edge server (ES) is responsible for computing all heavy cryptographic operations and uploading massive device data to a cloud. Nevertheless, the ES and the cloud may be attacked artificially in reality. Accordingly, to check the correctness of the ES data and the integrity of the cloud data, we design a lightweight dual auditing scheme, where the first adopts the aggregation of data blocks and signatures to check the integrity of the cloud data, and the second constructs an efficient Zero-Knowledge (ZK) proof to verify the correctness of the ES data in a lightweight way. To prevent collusion attacks, the legitimacy of the data user is authenticated by a smart contract. The simulation experiment shows that our multiple auditing scheme is relatively efficient.