Abstract
M&A plays an increasingly important role in the present business environment. Companies usually conduct M&A to pursue complementarity from other companies for preserving and extending their competitive advantages. A central and critical factor of the success of M&A is the appropriate selection of one or some target(s). However, existing studies reveal some limitations, such as the absence of technological variable and the exclusion of the characteristics of the acquirer in learning the M&A prediction model. In response to these limitations, we propose an M&A prediction technique which not only encompasses technological and innovative variables as prediction indictors but also takes both acquirer and candidate target into consideration when building an M&A prediction model. Ensemble learning algorithm and forty-three technological variables derived from patent analysis are applied to learn the M&A prediction model. Moreover, we collect two sets of real-world M&A cases to evaluate the proposed patent-analysis-based M&A prediction technique. The evaluation results are encouraging and will serve as a basis for future studies.