本文採用網絡多層級(廠商、成對與網絡層級)的觀點，深入探討不同層級的網絡結構㈵性對探索型與應用型創新之影響，強調網絡內的廠商可利用不同層級的網絡結構特質累積及獲取網絡資源與知識，以提高創新績效。本文以高科技產業及汽車產業之製造商為研究對象，發現網絡中心與及結構自主性會正向影響廠商的探索型創新；廠商與網絡成員之間多重性關係會正向影響廠商的應用型創新；當網絡穩定性提高時，將強化多重性關係對應用型創新的影響。此外，當網絡密度提高時，會增加中心性與結構自主性對探索型創新。

In this paper, we adopt multilevel network approach (i.e. firm-level,
dyadic-level and network-level) to discuss the influences of different network-level properties on exploitative and explorative innovations. Moreover, we emphasize that firms can utilize different network-level properties to obtain and accumulate resources and knowledge from the network in which they are embedded for enhancing innovation performances. We use manufacturers from the high-technology and automobile manufacturing industries as research targets, and find that: (1) the network centrality and structural autonomy positively influence a firm’s explorative innovation; (2) the relationship multiplexity positively influences a firm’s exploitative innovation; (3) an increase in network stability will
strengthen the positive relationship between relationship multiplexity and
exploitative innovation; (4) an increase in network density will strengthen the positive relationship between structural autonomy and exploration innovation and the positive relationship between multiplexity and exploitative innovation.