Absrtact: When you launch a public chip, it is not always possible to meet the established fund-raising goals. Here are two questions: what projects attract investment? How do investors find projects that are consistent with their goals? Recently, a team table at Cambridge University
When you launch an audience, it is not always possible to meet the established funding targets. Here are two questions:
What projects attract investment?
How do investors find projects that are consistent with their goals?
A team at Cambridge University recently said they were trying to answer both questions by digging into existing data. The algorithm, based on data from July 2013 to October on the Kickstarter of the famous platform, has more than 1000 projects and nearly 80,000 investors.
At the same time, they collected information on Twitter about "Kickstarter" during this period, with a total of more than 70,000, and if the message comes with a link to the project, match it to the project. This makes it possible to keep track of what each project has to say and what the people involved in the investment are.
They divide the investors into two types: the occasional investor, which accounts for about half of the time, and the one who invests more than 30 projects, which accounts for about 11%.
They found:
1. People who invest regularly have clear standards, they like this kind of project: good management, commitment to the goals of the broad, can be spread around the world, rapid development and so on, while the occasional investment in the standard is not the same, more inclined to contribute to the arts-related projects, and suspected that these people are affected by family or friends to donate.
2. The project sponsors of many Facebook friends are more likely to attract occasional investors, while Facebook's friends are not many, and they are more likely to attract regular investors. And those who often support the projects are similar to the professional investors, and the occasional supporter is more like charity.
Based on their own set of machine learning algorithm, the best case, the accuracy of the prediction of the match has reached 84%.
The team is set up to build a website that helps the public to find the right potential investors on Twitter. Of course, this idea can be used in turn, is to help investors find possible projects to vote. However, the specific release time is not clear.