Investigation on complex networks in software engineering

According to the statistic data reported by a famous international investigation company, the successful rate of large projects in software engineering is lower down sixteen percent. In the field of software engineering, human has proposed a huge number of ideas, theories, and methods since the technical term "software engineering" appeared. And new idea, theory and method are keeping increasing.. But, to compare with other one, why software engineering has so lowest successful rate? Many outstanding researches, engineers strive half century more for it in this field, and the situation is still very dejected radically. In this paper, we attempt to analyze some causes by discussing the complex network theory. And then, we try to seek for some solution ways to raise the project successful rate in software engineering. The complex network theory is developed in recent years, it studies so called "small world phenomenon", and includes many interesting research results. And after our studies, we found there are many complex networks in a software project developing process, and at the same time, they can explain byusing the research results in the field of complex networks. We discover that many complex networks in a software engineering process are independent, not connected among them, and not communicated by each other. For instances, software version management, software development schedule management, finance management, manpower management, and so on. To follow construction of the human brain as a model, complex neuron network can be divided into a large number of mutual to be provided with complex connection sub networks. We are planning to find a new method to link forward enumerated many management activities as a complete managing network system. WE hope that our research result could increase and raise the successful rate for large software developing.