FACTORIE is a toolkit for deployable probabilistic modeling, implemented as a software library in Scala that provides its users with a succinct language for creating relational factor graphs, estimating parameters and performing inference.
I like how easy it was to use Factorie compared to other tools out there we've used before; also liked being able to create models using only our data set instead of having someone else do all this work up front - which can be very time consuming! The documentation could have been more thorough/in-depth (some things are not explained well at first). We were trying learn about Bayesian networks but didn't really understand them until after reading some tutorials online or watching videos fromβ¦
Pros
Easy interface
Simple process for building new model(s)
Ability make predictions in realtime without waiting hours upon end
I enjoy using Factorie to work through problems related to Bayesian networks or graphical models - it makes things so much easier than trying them out by hand (as if you could do this without some sort of computer!).
The documentation can be confusing at times but once an example works well enough my confidence grows! When starting off new projects there are often multiple ways/tools people use when modelling these sorts of relationships; however none seem particularly intuitive nor easy-toβ¦
Pros
It also gets ridiculously fast because all computations take place locally instead sending anything far away into memory first then back again