In this paper, we propose solutions to resolving the domain mismatch problem. Processing data of mismatched domains in multiple databases will result in uncertain data in various situation. In the past, these data were considered to have the same level of uncertainty. We modify the past model and classify the data uncertainty into multiple levels in accordance with the so-called domain mismatch hierarchy. Our solutions enhance the previously proposed approaches in the literature by allowing the user to specify the way in which the uncertain information is extracted. In this manner, the amount of uncertain information derived from the mismatched domains can be limited to that with lower uncertainty so as to improve the trustability of the answer. Also, a probabilistic approach is proposed which utilizes domain knowledge to find the optimal result.