過去社會網路分析的研究中,社會網路圖內各參與者之間是否存在具時間序列調控關係的並未深入被探討,此訊息對於社會網路的分析者而言是一項非常珍貴的訊息,金融交易市場供應鏈中股票價格存在著價格連動性,因此本研究將以視覺化方式呈現金融股票市場的社會網路圖,公司在龐大的網路上形成關係密切的社群,股票價格的漲跌具有時間序列的調控關係,利用Pearson相關係數將兩間公司股價進行數列相似度計算,並獲得正負關係,在以動態時間軸校準(DTW)計算兩者是否具有延遲,藉此探勘出社會網路內任兩者之間是否具調控關係。
The existence of time series regulatory correlation among actors of a social network has not been investigated in depth in previous social network analyses. Such information is actually valuable to social network analyzers, especially about the correlation of stock prices within a supply chain. Our study, therefore, aims to visualize the social network graph in stock markets. In the massive social network composed of companies, the rise and fall of their stock index are temporally correlated. We use Pearson correlation coefficient to calculate the association of two companies’ stock prices and obtain positive or negative correlation. Then we measure if there is time delay between them with Dynamic Time Warping, and thereby determine the existence of a regulatory relationship between any two actors of a social network.