- 學位論文
利用文字探勘於延伸目標在群眾募資成功之研究
Predicting Stretch Goal Success in Crowdfunding Campaigns Using Text-mining Techniques
摘要
最近，眾籌已經成為一種越來越流行的籌款方式。為了有效利用多餘的資金和剩餘的籌款期，開發了一種稱為伸展目標的機制。延伸目標是通過向支持者承諾新的獎勵或其他激勵措施而設定的超出創作者初始資金目標的擴展目標。儘管如此，許多研究已經驗證了眾籌活動初始目標的成功因素，但很少關注延伸目標的成功。本文首先利用方面情感分析和文本挖掘技術探討了拉伸目標影響眾籌績效的關鍵因素。為了填補研究空白，我們進一步開發了一個預測模型來預測眾籌結果，並確定與延伸目標成功顯著相關的特徵。數據是從 409 個眾籌活動中收集的，這些活動在 Kickstarter 上獲得了延伸目標設置。結果表明，延伸目標的數量、創作者支持的項目和評論中的情緒對設置延伸目標的籌款績效做出了重大貢獻。我們的研究結果進一步擴展了對延伸目標設置中特徵的理解，並有助於優化眾籌績效。
關鍵字
none並列摘要
Recently, crowdfunding has become an increasingly popular way of fundraising. A mechanism called the stretch goal was developed to effectively use excess funds and the remaining fundraising period. Stretch goals are extended targets set beyond a creator’s initial funding goal by promising backers new rewards or other incentives. Nevertheless, many studies have verified the success factors of the initial goals in crowdfunding campaigns, but little attention is paid to the success of the stretch goal. This paper first explores which key stretch goal factors affect crowdfunding performance using aspect-sentiment analysis and text-mining techniques. To fill the research gap, we further develop a predictive model to forecast the crowdfunding outcomes and identify the features significantly associated with the success of the stretch goal. Data were collected from 409 crowdfunding campaigns with stretch goal settings obtained on Kickstarter. Results demonstrate that the number of stretch goals, projects backed by creators, and sentiments in comments have contributed to fundraising performance with stretch goal settings. Our findings further extend the understanding of features in stretch goal settings and help to optimize crowdfunding performance.