Finding gene relations has become important in research, since finding relations could assist biologists in finding a genes functionality. This article describes our proposal to combine microarray data and literature to find the relations among genes. The proposed method tries emphasizes the combined use of microarray data and literature rather than microarray data alone. Currently, many scholars use clustering algorithms to analyze microarray data, but these algorithms can find only the same expression mode, not the transcriptional relation between genes. Moreover, most traditional approaches involve all-against-all comparisons that are time-consuming. To reduce the comparison time and to find more relations in a microarray, we propose a method to expand microarray data and use association-rule algorithms to find all possible rules first. With its literature text mining, our method can be used to select the most suitable rules. Under such circumstances, the suitable gene group is selected and the gene comparison frequency is reduced sharply. Finally, we can then apply dynamic Bayesian network (DBN) to find the genes interaction. Unlike other techniques, this method not only reduces the comparison complexity but also reveals more mutual interactions among genes.