Abstract: 【Purpose/significance】This paper uses multi-source data to disambiguate the names of authors of scientific and technologi⁃ cal documents, so that the authors and scientific and technological documents have a one-to-one correspondence.【Method/process】In this paper, firstly, the collected multi-source data are preprocessed to form a duplicate name data set composed of the same name au⁃ thor literature to be resolved. Then, the academic circle is constructed through cooperation to find ambiguity. Finally, the ambiguity is eliminated through institutions and fields.【Result/conclusion】The experiment collected the literature data of six different subjects, such as education at all levels, automation and computer technology, information and knowledge dissemination, mathematical science and chemistry, radio electronics, Chinese medicine and so on. The rule-based disambiguation proposed in this paper has good disam⁃ biguation effect. Through multi-source data fusion, multi-target disambiguation of mechanism and domain, high disambiguation effect can be achieved.【Innovation/limitation】The problem of disambiguation in the same organization and field is solved, and the incremen⁃ tal problem is considered to build a complete disambiguation model.