投資決策輔助系統在金融操作上之應用──以台指期貨為例

Application of Investment Decision Making Assistant System in Financial Operation-Taking Taiwan Index Futures as an Example

期貨市場投資人不需研究個股，只需掌握趨勢判斷漲跌，就可決定買賣或觀望，有著以小博大優勢來獲取高報酬，但也伴隨著高風險，一旦操作失當，就產生巨大損失。本研究以本體論為基礎，建構消息面概念詞之知識庫與權重表，並透過資訊擷取與語意網之推論技術，過濾當日網路財經新聞內容，擷取符合知識庫之概念詞賦予權重。最後以2011年的歷史資料作為系統訓練樣本，得出適當參數值，並對2012年的期貨指數進行測試，透過模擬交易驗證系統之可用性，以得知系統功能績效，藉此建立一套投資期貨交易時的決策輔助系統之基礎開發模式。

It is unnecessary for investors in futures market to study individual stocks but judge the trends of ups and downs so that they can decide to buy or wait. It can get high remuneration with small money. However, it is of high risk. Once you make wrong operation, you will lose greatly. Based on ontology, this study builds information concept knowledge database and weight table and deduces technology by information collection and semanteme network. By filtering financial news information, it collects concepts weight fit for knowledge database. Finally, taking historical materials as system training sample in 2011, this study obtained proper parameter value and tested 2012 futures index. By mock trading, it verified usability of the system to learn its functional performance so as to set up a basic development model of decision making assistant system for futures investment.