台灣新創產業聲量趨勢分析-以Meet創業小聚網站為例

近年來台灣為積極推動新創環境的發展，投入相當多的資源。隨著創業熱潮興起，學者陸續提出「創業生態系統」（Entrepreneurial Ecosystem）的新興概念，引用自然科學的生態觀點，嘗試提出新的框架來理解此創業現象。因此本研究以「創業生態系統」此概念為出發點，將其中關鍵行動者之一：「網絡社群」為研究主軸，以「Meet創業小聚」網站為焦點案例，用文字探勘方式，觀察2010~2021年台灣七項重要新創產業提到的關鍵字聲量，並爬梳各產業的年度大事記，分析其聲量高低峰是否與當年的重大事件如政府發布有利於該產業創業環境的政策或因應國際上潮流而迅速發展等事件相符合，有助於了解台灣新創產業實際上的發展軌跡。研究發現主要有四，首先，從七項新創產業（「電子商務」、「區塊鏈」、「人工智慧」、「物聯網」、「交通科技」、「健康醫療」、「綠能」），觀察出自2010年至2021年的聲量曲線，可以得到兩種結果分別為：Meet創業小聚網站上的新創產業聲量與Google trend產業聲量曲線相似、或兩者曲線不相似。曲線相似者為：區塊鏈、人工智慧、物聯網這三類產業；曲線不相似者有電子商務、交通科技、健康醫療、綠能等這四類產業。其次，發現曲線變化與新創產業年度的獲投金額與件數有高度相關。第三，經由討論聲量高低，對17項新創產業做年度排名，得出年度產業排行榜。第四，從17項新創產業的聲量曲線可觀察出其高峰年度與該年產業的重大事件有關，證實可從新創產業的關鍵字聲量圖，找出高低峰時期，同時爬梳當年是否受一些事件影響，而若發現聲量與爬梳的大事記不一致，則可以進一步推論可能導致的因素，探討之間的落差。

關鍵字

台灣新創；文字探勘；創業生態系統；新創產業；產業趨勢；討論聲量

並列摘要

In recent years, Taiwan has invested considerable resources in startup development. With the rise of the entrepreneurship, scholars have successively put forward the emerging concept of "Entrepreneurial Ecosystem", which cites the ecological viewpoints of natural science and try to propose a new framework to understand this entrepreneurial phenomenon. Therefore, this study takes the concept of "Entrepreneurial Ecosystem" and takes one of the key actors: "Online Community" as the main branch and use the "Meet Global" website as the focus case by text mining method to observe the volume of keywords in seven important startup industries in Taiwan of the year 2010~202. Also review the annual event of each industry and analyze whether the peaks are related to important events such as the government's announcement of that year, or in order to international trends. It is helpful to understand the development fact of Taiwan's start-up industries. There are four main research findings. First, Observe the volume curve from 2010 to 2021 in seven industries (“E-commerce”, “Blockchain”, “Artificial Intelligence”, “Internet of things”, “Transportation Technology”, “Healthcare”, “Green energy”). Two results can be obtained: the volume of the startup industry on the Meet Global website is similar to the volume curve of the Google trend, or the two curves are not similar. Those similar curves are: blockchain, artificial intelligence, and the Internet of Things; those different curves are e-commerce, transportation technology, health care, and green energy. Second, it is found that the curve change is highly correlated with the annual investment fund amount and number of new ventures. Third, an annual ranking of 17 new industries was made through the volume of discussion. Fourth, through the volume curves of 17 new industries, it can be observed that the peak year is related to the major events of the industry in that year. if it is found that the voice of discussion is inconsistent with the annual event, it is possible to further infer the possible factors and explore the gap.