Study From Social Software, Web2.0 To Complex Adaptive Information System

With the rapid development of information technology, information systemevolves from low-level to high-level, simple to complex, close to open, and isolatedto cooperative. The components of information system become more dependent, thecoupling degree among components gets lower, while the components interact andcollaborate more flexibly. In a word, information system show more characters ofcomplex system than before. It is especially truth when social network software(SNS)and web2.0 are concerned.Different from traditional software and web information system, SNS andweb2.0 have some typical characters as follows: (1) They use moreparticipative-architecture and open-architecture;(2) There are plenty of nonlinearand self-organized mechanism in system, these mechanism can make the systemadjust its function and structure continually to adapt to the changing environment;(3)The interactions and inter-operations among these systems become very frequent andcomplex than ordinary ones, the complex relation of the interactions andinter-operations facilitate a dynamic complex network coming into being,Thecomplex network, which is quite similar to ecology network;(4) Social networkanalysis has been applied to information system design as embedded algorithms. Inthese systems, spontaneous cooperation among users allows many kinds of socialnetworks to grow up from the bottom.Equipped with these new characters, SNS and web2.0 are far more complexthan common information system, and their dynamic and complex characters aremore close to that of social system and ecosystem. When information system switchto complex system, many new problems and challenges is coming related to designand engineering study such system, for examples: (1) How to understand andconsider the dynamic mechanism of information system;(2) How to design adaptinformation system while adaptability is difficult to define and assess;(3) Thedynamic behavior of complex information system are difficult to be ruled, the futureprospect of complex information system are difficult to be predicted;(4) How torecognize the development law of information system and information technologywhen view them as a complex ecology network;(5) The dynamic complexity andadaptive evolvement of system make repeated test become nonfeasible (the stabilityof system function and performance is a necessary presumption for usability test andperformance contrasting test).In order to study and design complex information system effectively, thisdissertation proposes a new information system paradigm based on the theory ofComplex Adaptive Systems (CAS), and it is named complex adaptive informationsystem (CAIS). The paradigm of CAIS includes the basic idea and philosophy of thecomplex information system, a conceptual IS model which inherit from CAS directly,a research frame which guides to upgrade a traditional design IS to complex adaptiveinformation, and as well as some general design principles for CAIS. Then we dosome detailed researches under frame of CAIS to solve the above-mentionedproblems in complex information system.The main content of this study can be summarized as follows:Different elements of CAIS constitute complex networks with classified nodes.Analyzing and clustering these networks helps facilitate data and components withcommon characters coming together and person sharing common interests or tasksflocking together, which can be demonstrated by (1). Having components with innerpertinence related each other favors to optimize combination of function and contentsearching , (2). Having users with similar objective and interesting related each otherhelps build up cooperation among them. Based on this application background, thispaper puts forward some analytic algorithms of complex network constituted by twoor three kinds of nodes(including algorithms for splitting 2-mode network into two1-mode networks and strategies for analyzing 3-mode network, etc).Finally, ageneral frame for embedding these algorithms in CAIS are provided to improve theadaptability of system.CAISs possess much more dynamic complexity than ordinary informationsystems. For a fresh running CAIS,it is possible that there are many functions andcharacters of IS not presented for new users but for old users, not presented whenthere are only a few users but presented when there are enough users. The wholesystem becomes more adaptive and more convenient to use. By that means, systemfunction and performance are changeable at different time for different person thatdepend not only on given person usage of system but others in and out of the samesystem, which make it very difficult to layout and design system function andappraise the performance of system optimization. Thus, conventional prototype thatis suit to describe static system can not model such dynamic complex system, andtesting repeatedly by conventional test method is also impossible. Agent-basedmodeling(ABM) has been proved to be a effective method to study the dynamicstructure and mechanism of emergence for complex system. In this research weintroduce ABM into assistant design of CAIS and built 2 models to solve the abovetwo problems.Along with more and more cross-system mixture appearing,the coupling degreeamong information systems is becoming higher(contrasting the decline trend amongcomponents within a system ). Whether a new pattern succeed depends not only onits own design, but its adaptability to other information system(collaborationcapability, inter-operation ability, integration ability and amalgamation ability withother information system,etc). We define information system ecologies to besystems of CAISs in a particular network environment. In information systemsecologies,one CAIS is an agent of a larger CAS. This paper argues applyinganalytical measure of ecology to help conceive and reason new project, some typicalmodels and developing disciplinarian deriving from new complex informationsystem are summarized as well.The main contribution of this paper is: in allusion to more and more complicatedand diversified SNS and Internet constructed by more perplexing Web informationsystem, theories and measures of system science and complexity research areroundly introduced into information system research for the first time. Theapplication of system thinking, theory, method to information system analysis anddesign makes up insufficiency of theory research for such new born informationsystem as SNS and Web2.0, extends the application scope of system theory andcomplex system research as well. In addition, some new problems of applyingtraditional research method to new research field are found and brought about,during the process of solving such problems, some system research methodsthemselves are reinforced and improved.