c) A yukawa interaction is a type of interaction between a scalar field and a fermion field, i.e.
http://en.wikipedia.org/wiki/Yukawa_interaction,
But I guess you do not know what the difference between scalar fields and fermion fields is, so forget that. As far as the electroweak Lagrangian goes, the thing to know is that Yukawa interactions are the ones that act like mass terms for the fermion fields, i.e. the quarks and leptons.
e.g.
y\Psi\phi\Psi
would be a Yukawa interaction between scalar field \phi and fermion field \Psi (with some indices and conjugates and such left out). The 'y' is the coupling strength of the interaction. The 'matching' mass term describing a fermion field with mass m would be
m\Psi\Psi
So if the scalar field (Higgs field) was to stop being a dynamical field, and just adopt a constant value \phi_{0}, the Yukawa term would look exactly the same as the mass term, with m=y\phi_{0}. And that's how you can generate mass terms without just sticking them into your Lagrangian (which you aren't allowed to do in the Standard Model because they break gauge invariance)
edit: Ok I actually read that previous response now and have this to add:
a) Perhaps you are looking at a page something like this:
http://en.wikipedia.org/wiki/Standard_Model#Lagrangian. The W and B and G in that Lagrangian are field strength tensors, not the fields themselves (although they are functions of the fields). When combined in the Lagrangian in that way they describe the dynamics of the fields they relate to; they more or less generate Maxwell's equations and equivalent things for the other forces. You'll have to learn some special relativity to appreciate those indices properly.
b) The word 'covariant derivative' is thrown around in different contexts. The previous response talks a little bit about it in terms of a curved manifold, which in general relativity would mean curved spacetime. It is how you do derivatives in curved space.
However, you are asking about the electroweak Lagrangian, and in this context it means something a bit different. The Standard Model is constructed in flat space, so there are no curved space covariant derivatives to worry about. Instead, there are what are sometimes called 'gauge covariant derivatives', which are how you do derivatives while preserving the gauge symmetries of the model. The two usages are really manifestations of the same underlying math but things are complicated enough without worrying about reformulating the standard model in geometric terms.
Have a read of the wiki:
http://en.wikipedia.org/wiki/Gauge_covariant_derivative
I am just trying to throw some conceptual ideas out there, because you have a lot more to learn before you can really appreciate what is going on. I do too though, so don't be discouraged if it all seems like gibberish. It will continue to seem like gibberish for a long time yet :).