確認影響電影群眾募資專案之關鍵成功因素

Identifying the Key Successful Factors of Movie Crowdfunding Projects

莊穎融(YING-JUNG CHUANG)

指導教授 ： 陳隆昇

朝陽科技大學/資訊學院/資訊管理系/碩士(2020年)

摘要

群眾募資（Crowdfunding）已成為許多新創企業獲得資金的重要管道之一。但低成功率一直是一個關鍵問題。因此，如何提高籌款專案的成功率是所有籌款活動的主要關注之一。研究指出專案描述會影響群眾募資專案的成功，專案描述的內容將影響支持者的決定，但仍未有研究深入探討專案描述文字應包含哪些要素，會直接影響專案成功。此外，有研究指出專案參與者的評論情緒會影響群眾募資專案的成功。根據近年群眾募資研究，分別在音樂領域、體育領域等研究對專案募資成功有積極的關鍵因素，但尚未針對電影做探討。因此本研究將研究於電影群眾募資專案的成功影響因素，以專案描述文字與參與募資者的評論為主要研究主體，取代以往文獻使用於問卷調查做為資料收集的來源。因社群評論的影響力日增，所以我們將使用資料探勘和文字探勘方法來分析收集的專案文字描述與評論文字資料，希望改善問卷調查的抽樣誤差的問題。本研究將利用特徵選取方法包含決策樹(Decision tree, DT)、最小絕對收縮和選擇算子（Least Absolute Shrinkage and Selection Operator, LASSO）和倒傳遞神經網路（Back Propagation Neural Network, BPN）來找到影響電影群眾募資專案成功的重要因素，並使用支持向量機（Support Vector Machine, SVM）來評估候選之特徵子集，預期實驗結果可幫助籌款者提高電影類群眾募資專案的成功率。

關鍵字

群眾募資 ； 特徵選取 ； 決策樹 ； 最小絕對收縮和選擇算子 ； 倒傳遞神經網路 ； 績效評估

並列摘要

Crowdfunding has become one of the important channels for many startups to obtain funds. But low success rates have always been a key issue. Therefore, how to improve the success rate of fundraising projects is one of the main concerns of all fundraising activities. The study pointed out that the project description will affect the success of the crowdfunding project, and the content of the project description will affect the supporter's decision. There has been no research to discuss in depth what elements should be included in the project description text, which will directly affect the success of the project. In addition, some studies have pointed out that the sentiment of project participants' comments will affect the success of crowdfunding projects. According to the research on crowdfunding in recent years, studies in the music and sports fields have positive and critical factors for the success of project fundraising, but have not yet discussed the film. Therefore, this study will study the influencing factors of the success of the film crowdfunding project, using the project description and the comments of the fundraisers as the main research subjects, replacing the previous literature used in the questionnaire survey as the source of data collection. Due to the increasing influence of community comments, we will use data exploration and text exploration methods to analyze the collected project text description and comment text data, hoping to improve the sampling error of the questionnaire survey. This study will use feature selection methods including Decision Tree (DT), Least Absolute Shrinkage and Selection Operator (LASSO) and Back Propagation Neural Network (BPN) to find Important factors that affect the success of film crowdfunding projects, and use Support Vector Machine (SVM) to evaluate the candidate feature subset. The expected experimental results can help fundraisers improve the success rate of film crowdfunding projects.

並列關鍵字

Crowdfunding ； Feature Selection ； Decision tree ； Least Absolute Shrinkage and Selection Operator ； Back Propagation Network ； Support Vector Machine

參考文獻

1. Alamsyah, A., Nugroho, T.B.A., 2018, Predictive modelling for startup and investor relationship based on crowdfunding platform data,” IOP Conf. Series: Journal of Physics: Conf. Series 971.

2. Agrawal, A., Catalini, C., Goldfarb, A., 2015, “Crowdfunding: geography, social networks and the timing of investment decisions,” J. Econ. Manag. Strateg, Volume24, Issue2, pp. 253–274.

3. Arena, M., Bengo, I., Calderini, M., Chiodo, V., 2018, “Unlocking finance for social tech start-ups: is there a new opportunity space?,” Technological Forecasting & Social Change 127, pp.154–165.

4. Aleksina, A., Akulenka, S., Lublóy, A., 2019, “Success factors of crowdfunding campaigns in medical research: perceptions and reality,” Volume 24, pp. 1413-1420.

5. Ahlers, G.K.C., Cumming, D., Guenther, C., Schweizer, D., 2015, “Signaling in Equity Crowdfunding,” Entrepreneurship Theory and Practice 39, pp. 955–980.