探討財經新聞與金融趨勢的相關性 在這個資本主義的時代，市場上有許許多多金融相關的產品可以做選購，但產品的玩法過於複雜，且市場的消息也太過紊亂，又時常充斥者假消息，對於一般人實在很難做分辨並且理解出，新聞與金融產品之間的關係，本論文研究透過深度學習的方法，和對於網路上的資訊的探勘，以及使用對文本的情緒分析，在文章的分析階段結合時下流行的神經網路，來幫助使用者分析出哪些才是財經新聞內必須表達的重要意義，找出定義新聞關鍵詞的幾種三種想法，必且驗證出深度學習的方法在分析股市新聞的1/3/5/10/15 天候的漲跌是有幫助的，並且發現影響最劇烈為離新聞發布日最靠近的日期，希望可以讓投資人得到可用的參考資訊，並且可以證明財經網路和新聞漲跌的相關性，以幫助投資人得到有用的資訊，以降低投資風險。 ;The relevance between financial news and financial trends
In Capitalism, there is a lot of financial products could be chosen for earning more money, however, that product is too complex and the financial news is too messy for a customer to invest it. For an ordinary consumer, is hard to recognize the content of the news what the reporter really wants to tell to the reader. In this paper we want to figure the out the relevance between news and financial trend by training a model by deep learning method with the key of news content and the data mining at the internet information. With the sentiment analysis we thinking of put the key word at news content by 3 different ideas. Verified the idea for Deep learning’s prediction function on ups and down of 1/3/5/10/15 days latter stock’s price. Hope it can help customer to figure it out when is a best time to invest a product and whether the financial is effect by the news and prove the relevance between financial news and financial trends.
