Open innovation model is the best choice for the firms that cannot afford R&D costs but intent to continue playing the innovation game. This model offers to any firm the possibility to have companies spread worldwide and in all research fields as partners in R&D. However, the possible partnership can be restricted to the manager's know-who. Patent documents can be the source of rich information about technical development and innovation from a huge amount of firms. Search through all these daily created documents is a cumbersome task that technology managers cannot afford. This paper aims to introduce an automated model to seek matching firms' R&D using data mining techniques applied into patent documents database. The methodology considers the search for patent documents from possible partners and these data treatment through the association technique among IPC fields. An evaluation system was implemented and a sample experiment was made. The results reached are patterns of technological knowledge interdependence that can be used to evaluate four possible types of partnership.