As Jake, Sean and Ted have already described the advantages of joining Mahout from a contributors view, I’d like to add some practical considerations about the expectations potential end-users of your open-source code will have.
These “requirements” are not difficult to achieve but extremely time-consuming for a one-man-project, so I’d regard them as reasons to join a multipeople project:
A mailing list where questions are answered within a day or two.
A single person will be on vacation, starting a new job or be frustrated with life every once a while and will therefore not be able or not willing to answer questions in time.
In a community-driven project that’s perfectly ok because the people involved will rotate with activity on the mailinglist and there’s no problem with one contributor bowing out from time to time.
An extensive documentation and ideally a book to get you started.
This is even very hard to achieve in an Apache project and is seen very seldom in one-man projects from my experience where the developer has a regular job besides his/her open source work.
A release available as maven artifact.
I regularly find very cool software that does not provide maven artifacts and is therefore very difficult to use in professional projects.
A somewhat mature, community-driven project will already have a working release process.
The guarantee that the project will last longer than a single developers enthusiasm.
This is pretty much a no-brainer.
Building a big system on top of a library that could be abandoned by it’s only developer is a show-stopper for a lot of potential users of your code.
More Resources on Why should one contribute to Mahout instead of starting a new open source project:
Read more about Why should one contribute to Mahout instead of starting a new open source project: