The conventional network multi-source heterogeneous information integration method mainly uses the Arcgis Engine multi-functional development package to read and stack heterogeneous information, but this method is vulnerable to the impact of multi window data linkage, resulting in a low degree of information integration correlation. Therefore, a method of integrating multi source and heterogeneous information on the Internet is designed for hybrid English teaching. This paper constructs a network multi-source heterogeneous information integration ontology for hybrid English teaching, carries out the RDF integration and transformation of multi-source heterogeneous information for hybrid English teaching, and designs the network multi-source heterogeneous text classification information integration algorithm to complete the integration of network multi-source heterogeneous information. The experimental results show that the designed network multi-source heterogeneous information integration method for blended English teaching has an information correlation degree closer to 1 after integration, proving that the integration effect of the designed multi-source heterogeneous information integration method is good, reliable, and has certain application value. It has made a certain contribution to improving the utilization rate of blended English teaching resources and optimizing the blended English teaching mode.