The invention provides a data extraction method and device and a storage medium, data extraction is carried out based on a complex feature adaptive matching method and a variable sliding window method, the method comprises the steps of firstly, obtaining massive data from multiple data sources and then clustering, and setting a unique data label for each category of clustered data; and judging whether the category data is data of a known type or not according to the data label, if so, carrying out feature extraction on the category data by using a complex feature adaptive matching method, andif not, carrying out feature extraction on the category data by using a variable sliding window method, and storing a result in a database. Valuable information in known types of data is accurately and quickly extracted through a complex feature adaptive matching algorithm, feature cyclic comparison is performed on unknown types of original data through a variable sliding window algorithm, the missing extraction ratio of the valuable data is reduced, and the big data analysis and core clue quick positioning capacity is improved.