本研究以建構一個以社會網路分析法為基礎的股市投資決策系統為最終目的,本研究以TEJ資料庫做為股市資料來源並以外資券商作為主要研究對象來建構社會網路,並先對原始資料進行核心-邊緣分析,以找出存在於股市交易社會網路中具有高度關聯性的股票或券商,接著使用不同的節點以及關係定義來建構出不同的社會網路來進行分析。本研究先找出具有高度績效的券商或股票並建構成社會網路,再透過由核心-邊緣分析來分析這些社會網路來得到所需要的資訊。透過過去的資料來測試推薦的結果,最後綜合分析的結果來呈現出可參考之資訊,而使用本研究所提出的方法對過去之歷史資料進行實測後發現確實可以產生可用之資訊。本研究之決策支援系統將採用有效的社會網路分析法來建構其中的模式管理子系統部分,以支援投資者的決策過程,幫助投資者進行更佳的投資決策,降低投資風險。
In this study, the research objective is to implement a decision support system for stock investment based on social network analysis. This study uses TEJ database as a data source and foreign brokerages as the main objects to construct the social networks. The social networks are analyzed by Core-Periphery Model to find out the stock or securities with a high correlation, then filters social networks by different definition of nodes and tides for further analysis. The system uses past market data to validate the results of recommendation, then sums up analysis results for more information to support the decision system. The outcome of the proposed methods is tested and shows promising recommendation. The decision support system in this study uses social network analysis to build the model of management subsystem for helping and supporting the process of investors to make decision, reduce the risk and have the better support information while doing investment.