新聞除了是投資人用來得知上市公司的相關資訊的主要媒介，亦是屬於會影響股價顯著變化的非結構化因素。新聞事件的本身除了會影響投資人在決定其投資策略的重要參考依據外，同時也隱藏著具有影響股價變化的可能性。是以如何在眾多新聞面項中，找出提升投資績效的通用新聞知識結構是相當有意義且重要的研究。 本研究提出在建構出影響股價之企業新聞的知識本體模型。並利用此知識本體模型，使用資料探勘技術將新聞文件處理過後，利用本體中企業屬性分類新聞資料，使用關聯規則找出企業每個屬性項目的關鍵詞組。最後將該模型下之關鍵詞組結構，用來驗證此結構與市場上股價的變化之關係。 實驗結果找出了企業知識本體跟企業價值變化的關係，利用此關係可以去針對短期內企業價值變化在股票上的預測分析，作為投資人投資時的決策參考。 News is in addition to the investors used to know the relevant information of listed companies, the main medium, is also a significant change in stock prices will affect the unstructured factors. In addition to the news itself will affect the investors in determining their investment strategy based on an important reference, but also hidden with the impact of the possibility of changes in stock prices. It is quite meaningful and important to find out how to find out the general news knowledge structure that enhances investment performance in many news items. This study proposes to construct a knowledge ontology model of corporate news that affects stock prices. And use this knowledge ontology model, the use of data mining technology to deal with the news file, the use of ontology in the enterprise attribute classification news materials, the use of association rules to find the enterprise each attribute project keyword group. Finally, the keyword group structure under the model is used to verify the relationship between the structure and the change of the stock price in the market. The results of the experiment find out the relationship between the enterprise knowledge ontology and the change of the enterprise value. The relationship can be used to analyze the forecast of the enterprise value in the short term, as the decision reference for the investor''s investment.
