Abstract: Effective technology search is a premise for enterprises to develop new technologies and to further seek commercial value in the market, thus being necessary for enterprises to meet their technology needs especially in highly competitive industries. Although the explosive growth of knowledge and the extension of network provide much more space and path for technology search, the aggravation of market competition bring more risks to the technology search. Facing the limitation of technology search space and path as well as the constraints of internal resource and capabilities，enterprises have to constantly adjust and optimize their search strategies to save more searching costs and improve searching efficiency. This paper aims to reveal the efficient search strategies.The existing research of technology search has experienced an expansion from one dimension to two dimensions. At first, scholars mainly focused on the scope of technical search and divided technical search into local search and non-local search, internal search and external search or exploratory search and exploitation search according to cognitive boundary, organizational boundary, geographical boundary or time boundary. And then, some scholars put forward the concepts of search depth and search breadth and expanded the research of technical search to two-dimensional space. After that, some scholars found that technology search is not a discrete individual behavior and tried to use the theories of system science and ecological science to study the technology search behavior of enterprises. However, most studies ignore the analysis of technology value and only focus on technology search breadth and search depth, while those few scholars who pay attention to the value of technology in technology search have failed to construct a model with testable propositions. In addition, the existing empirical research focus too much on individual behavior and has not yet put enterprises′ technology search into the industry competition.Therefore, based on the existing research of technology search, combining the adaptive landscape theory and the view of innovation recombination, we revised the NK model of Kauffman (1995) and made it more suitable for technology search. As the behavior of enterprises to enhance their core technological competitiveness to adapt to the external market competition is similar to that of organisms to change their own genetic characteristics to adapt to the external environmental changes, this paper extracts the “genome”-technology position based on the IPC (international patent classification) combination of each patent and supposes that the greater the difference between two “genomes” means the greater the technical distance between the two corresponding technology positions, thus two adjacent technology positions have only one knowledge element (IPC) whose states are different in the two positions. In addition, every technology position is supposed to have specific value which is related to not only every knowledge element (IPC) but also the coupling relationships among those knowledge elements (IPC). Taking technology position as the basic unit, this paper constructs a three-dimensional technical landscape map which presents the distance among those technology positions and the value of each technology position. And based on this map, we analyzes the value characteristics and interrelations of the technology positions occupied by enterprises by using the revised NK model and put forwards some testable propositions about the search strategies for enterprises to search for high-value technologies more effectively. The empirical testing section is to prove the propositions derived from the NK model. In this section, pharmaceutical manufacturing enterprises above the designated size in China that have new product output in 2006-2007 and 2009-2010 are used as the sample to empirically analyze the propositions. We collected information of the enterprises from the Industrial Enterprise Database and that of the patents from the Innojoy Database. The value of one technology position is measured by the sum of the output value of new products created by the patents in this position. The empirical methods include both regression analysis and descriptive statistical analysis of cases. First, regression analysis is used to explore whether there is a significant correlation between the values of two adjacent technology positions. Second, statistical analysis of eight representative enterprise cases which are the tops in the output value of new products is used to verify the effectiveness of the strategies. We investigate the evolution path and value of these enterprises′ technological positions on the technology landscape map in the pharmaceutical industry.The results of this research show that: The values of two adjacent technology positions are related positively. In order to search for high-value technology more effectively and win more market share, enterprises should adopt different search strategies according to their different existing technical resources. For the high-value technology positions they have already occupied, enterprises should search new technology on the adjacent technology positions of these high-value technology positions, so as to occupy more high-value technology positions. For the low-value technology positions they have occupied, enterprises should expand their search space and path to improve the possibility of accessing high-value technology, most of the research and development efforts should be put into those non-adjacent technology positions far away from these existing low-value technology positions.The research significance of this paper are as follows: First, technologies with high market value contributes more to the promotion of the core competitiveness of enterprises, this paper provides guidance for enterprises to effectively search for high-value technologies based on their internal technology base. Second, technology position is a relative new basic unit to measure the technical characteristics of enterprises from the perspective of recombinant innovation. This paper clearly defines the concept of technology position, which provides a tool for the study of the adaptive behavior of enterprises in response to external market changes, and the technology landscape of an enterprise or an industry can also be drawn through the evolution analysis of technology positions. Third, the research shows that the technology search is not an independent behavior of one single enterprise but is embedded in the complex market environment, thus being affected by the related enterprises in the whole industry. Therefore, the construction of fair market competition environment is of great importance to promote the efficient of enterprises′ technology search behaviors.

Key words: fitness landscape, technology search, recombination search, technology position, technology position value