Title:

Detecting ebook spam:
Automatic determination of ebook quality

Description:

Electronic books (ebooks) are swiftly becoming a profitable market. So
profitable, in fact, that spammers have taken note. For example,
Amazon's Kindle platform is currently a prime target
for spammers looking to make easy money. They grab texts from all over
the web, copying texts, make duplicate copies of a book with slight
tweaks to title and author, in short: a barrage of spam ebooks is
flooding the market. The goal: profit. Even if the books are sold
cheaply: every sale translates into money for the author.

The goal of this project is to develop and implement a method to
automatically determine the quality of an ebook, taking into account a
variety of indicators. The first step is to determine a set of quality
indicators for ebooks, such as "number of books published by this author
per week/month/year", "total number of books published by this author",
"is this content copied from somewhere", "is this book being offered
under different but similar titles", etc. Based on these indicators, a
weighing algorithm is developed, allowing for future extensions with
further indicators. Next, the method is implemented as a FireFox plugin.

The student is expected to have a background in:
- Information security
- Web languages and web programming (Javascript, PHP, HTML, CSS)


Contact:
Hugo Jonker: hugo.jonker@uni.lu