Currently there is such an increasing interest in discovering important patterns from graph data. A significant number of applications require effective and efficient manipulation of graph mining, such as being: (i) analysis of microarray data in bioinformatics, (ii) pattern discovery in a large graph representing a social network, (iii) analysis of transportation networks, (iv) community discovery in Web data. This paper concerned with subgraph discovery from weighted graph data that came from the educational context. The non-linear correlation technology was introduced and used in the mining process in the whole knowledge achieved. At last we have applied these methods in real course management datasets and found correspondent results for the educators.