The invention discloses a multivariate big data fusion method and system based on a BLOB (Binary Large OBject). The method comprises the following steps that: 1) on the basis of an RDF (Resource Description Framework) data model, creating an RDF-B data model, adopting a triple form by the RDF-B model to express the attribute of each resource and the attribute value of the attribute, wherein the attribute value comprises the attribute value of a BLOB type, and the attribute value of the BLOB type comprises data contents, length, abstracts and symbol information; and 2) generating a triple for the received data by the RDF-B data model, and storing the triple into a front-end storage system, wherein if the data is of a BLOB type, the RDF-B data model generates a tetrad< handle, length, abstract and symbol> according to the attribute value of the BLOB type in the triple corresponding to the data as the attribute value of the triple information of the data, then, the triple of the data is stored into the front-end storage system, and the data contents are stored into a rear-end storage system according to the handle.