現今社群網路發展蓬勃，提供了豐富的信息來源，可以進一步探討和分析預測未知的知識或未來事件。本論文旨在應用大數據工具收集社交網路有關於手機的討論信息，此信息將進一步的分析，作為預測未來手機銷售，並找出那一種網路議題與手機銷售量間之關係 實證結果顯示以MATLAB R2016b軟體中的曲線擬合(Curve Fitting)所建立預測模型，發現網路議題"推薦"與銷售量有其相關性。本研究認為當月網路上的相關討論提到"推薦"此一個關鍵字的文章數量越多，下一個月的銷售量也隨著提升。在這個月的手機上建議的數量一般可以用於預測下個月的銷售。這樣的結果是對於一般手機用戶或商販非常有用。 Social networks have provided rich sources of information which can be further explored and analyzed for predicting the unknown knowledge or future events. This thesis aims to apply the big data tool to collect the information, including the sales of the mobile phone and the number of discussions related to the mobile phone, from social networks. This information will be further analyzed to predict the future sales of the mobile phone, based on the relation between the sales and the number of discussions on the mobile phone. Finally, the MATLAB tool is applied to verify the results. Experimental results show that the ‘recommendation’ and ‘sales’ of the mobile phone exist strong relation. The number of recommendations on the mobile phone in this month can be generally used to predict the sales of next month. This results is useful for general mobile phone users or venders.
