Skip to content. | Skip to navigation

Informatik 5
Information Systems
Prof. Dr. M. Jarke
Personal tools
You are here: Home Theses “Your Project Sucks!” Prediciting Kickstarter project success based on crowdfunding language


Prof. Dr. M. Jarke
RWTH Aachen
Informatik 5
Ahornstr. 55
D-52056 Aachen
Tel +49/241/8021501
Fax +49/241/8022321

How to find us

Annual Reports





“Your Project Sucks!” Prediciting Kickstarter project success based on crowdfunding language

Thesis type
  • Bachelor
Status Running
Proposal on 11. May 2017 17:30
Proposal room Seminarraum I5
Add proposal to calendar vCal
Kickstarter is a crowdfunding website that collected more than $2.5bn in pledges by 11.4mn backers since 2009. Sentiments and Emotions in online communication can range from supportive to extremely negative when projects fail to deliver on their promise. In this project, we like to investigate how far the communication structure and sentiment in comments on Kickstarter project pages is predictive for a) successful project funding b) successful project delivery to funders.

Methods In the project a student has to analyze comments to successful and failed Kickstarter projects using recent Natural Language Processing Opensource tools (Stanford CoreNLP) and other services for detecting sentiment and emotion (IBM Watson AlchemyLanguage API). Using an outcome as a feature vector, the student has to train a model for predicting project success using, for example, a Tensorflow framework.

Data (Scraped) Data on comment threads on Kickstarter will be provided as .txt or .json. In total, we have about 1,500 projects (or more, if necessary) to analyze. These projects received roughly a total of $14m crowdfunding. How do comments look like? An Example of project comments on Kickstarter

Collaboration The project will be conducted in collaboration with economic data scientist at the Technology Entrepreneurship Group (TEN) at RWTH Aachen.
Document Actions