Genealogy data is a typical example for data fragmentation with massive, multiple, heterogeneous, and autonomous sources. Merging scattered genealogy data on the Internet into a comprehensive and accurate genealogy database through data fusion technologies, can be beneficial to knowledge mining and reasoning from genealogy data, and can provide users with implicit information such as surname origins, surname changes, and surname associations. Based on BigKE, a big data knowledge engineering model for big knowledge, this study proposes an FDF-HAO framework (fragmented data fusion with human intelligence, artificial intelligence, and organizational intelligence), describes the functionalities, key technologies, and problems to be solved of each layer in the framework, and verifies the validity of the data fusion framework by using genealogy data as an example. Finally, the challenges and opportunities of fragmented data fusion are also discussed.