時至今日，利用社群網路平台發表自我感受及想法成為趨勢，使得產品評論資訊更容易的被發表及傳播。本研究結合『未來學』、『意見探勘』、『情感分析』三個領域。本研究採用 Nunamaker (1991)等人所提出之系統發展研究法，旨在分析網路評論，整理出網友對於產品的期望，讓開發商更直接簡便的了解消費者的當下需求與未來趨勢。本研究的實驗對象為華碩公司所推出的手機「Padfone變形手機」、「Zenfone智慧型手機」兩種系列產品的評價，實驗資料集來自Mobile01小惡魔論壇的文章資料。文章日期蒐集範圍分別以該系列首次發表年度開始。整體準確率達89%。而未來詞與意見單元之修飾距離為8以內時，未來性意見準確率達60%。最後以圖表呈現產品的意見狀況，表達使用者對產品各部件的意見與期許。 Texting products usage experience and comments in social media community has been a tendency of most people new behavior. This research is to develop the model of combined with “future related wordings”, “opinion unit retrieval”, and “sentimental analysis” methods for exploring and analyzing text-oriented people opinions and comments. This system prototype adopting with Nunamaker(1991) system development method is to demonstrate the amount of positive and negative in time-frame curve of product opinions as well as to search the possible future related wordings for estimating the needs of product future design. This study is based on ASUS mobile products data samples to manipulate opinion unit mapping process and the result of this experiment is to find the accuracy rate 89% of matched opinion unit, the distance between “future related word” and opinion unit less than 8 words in most sentences, and reaching 60% corresponding to “future” definition. Using Graphic style describes user product opinions and product expectations as deducing to future related words in final conclusion of this research.
