Okay, so welcome everybody.

We talked almost all of the semester about the method of fragments and we actually looked

at four fragments.

This is what you finished, essentially, on Thursday.

For some value finished.

The aim of this lab session is essentially that not that we have a finished piece of software that works, but that you know how to complete this into a finished piece of software that works.

So, we've essentially covered the whole

area

up to a degree where you should be able to continue on the love.

In a sense that

you know what the fragment does.

And

you could look into the literature.

There are lots of papers that basically do exactly what we've been doing with fragment one to four.

So, you can define a fragment by giving a grammar, or more honestly often hinting at a grammar.

And then kind of look at certain things in isolation, giving a semantics construction and then possibly even detailing a logic.

And that's just first of all logic.

Well, our friends from computer science can do the inference that we need.

And those you could actually all take

convert

in the kinds of system we've shown you.

And that at the time didn't quite exist. That was essentially what you could do was you could take Prolog as the system and just implement extensions to Prolog.

That's essentially it.

So, doing context free grammars in Prolog is relatively simple.

You could just use Prolog predicates as parsing rules and so kind of cobble something like this together.

Of course, there were very severe scalability issues, even more severe than the ones we're facing in here.

Are there any questions about that?

We have a homework assignment that Frederick is uploading

to Stodon, I think he said, as we speak.

It's essentially about dealing with this Neo-Davidsonian event semantics.

And doing more generation on that. It's essentially doing a further completion on

fragment theory.

The next thing I would like to talk about with you is the exam.

We have to have one.

And I know all of you proclaim they don't need grades, which I think is not going to happen.

So we have to have an exam.

And I would like you to think about when you would like that.

I won't be here, so Frederick will do it essentially alone.

Well, we're going to design the exam together, but Frederick will proctor it.

And we're relatively open.

Any suggestions? We could do this essentially by section.

There's one possibility which is essentially directly after the semester.

I see heads shaking.

It could be directly before the next semester. So something like

early April.

Directly after the semester has the advantage that you still remember what was going on.

And don't have to relearn.

Directly before the next semester has the advantage that you have lots of time.

And you have the opportunity to relearn.

So what's it going to be? I don't care.