In recent years, digital publishing share in the information society more and more, readers of the growing demand for personalized service strong traditional culture, publishing services are currently unable to meet this demand, especially for the digital home needs. For family-oriented, personalized digital publishing needs, this paper uses the method of multi-dimensional indexing metadata tissue, data mining, knowledge classification management, natural language as the three dimensions of the dynamic association to create a database; mechanical matching personalized method of data retrieval, and can take the initiative to push; through the reorganization of the publishing and content transcoding method, the layout can keep the original format of the EPUB format document; using digital watermarking and digital encryption to protect the method of combining the author’s legitimate interests. In this article based on the development of family-oriented personal digital publishing services system, in practice, demonstrate the practicality of this technology.