I run the University of Southampton Open Data Service. http://data.southampton.ac.uk/
For the love of all that is holy, don't invent a new vocab -- there's plenty of solid work already out there.
This site is designed to be a model for other people to lift ideas from. We've used simple solutions which are easy for users to consume and all the tools are open source. We mostly use simple spreadsheets, then XSLT to produce RDF data. All of this is also open source. All of the tools used to generate each dataset are published with the dataset to make it really easy to either directly copy our approach, or use it as a starting point.
Picking good, simple, sustainable URIs is key. We use a domain to do this and only this: eg. http://id.southampton.ac.uk/building/59
People -- start with FOAF.
Person Roles -- I suggest you look at the "organogram" produced by the UK civil service.. this treats a role as a type of group which contains zero or more people. (multiple people can have the same role, and roles can exist where people are empty)
Publications -- we use bibo and dcterms for this. The trickiest part is neatly linking it back to people & parts of the organisation etc.
Meetings -- we don't currently model this, but you can just use the standard "event" vocab unless there's a reason not to: http://motools.sourceforge.net/event/event.html
Projects -- this is something we're just starting to consider, and my current plan is to use http://ontoware.org/swrc/
Research area; I don't have any useful ideas here, yet.
We are trying to document sensible starting patterns for this kind of task on openorg.ecs.soton.ac.uk -- which will be a constant work in progress. The goal is to produce data with enough commonality of structure that you can create apps which work with more than one organisation, and more easily merge information from other sites.
I've written a blog post demonstrating one tool working with (places) RDF data produced by Southampton, Oxford and Lincoln (although the Lincoln data server is offline right now) http://blogs.ecs.soton.ac.uk/data/2011/11/25/roi/
We have an RDF namespace called "openorg" for the additional vocab we've found we need to achieve our tasks. We try to keep this to a minimum and most things are described using existing vocabs.
As for building the site on top of the data, we are using a PHP Library called "Graphite" (which I'm the lead developer on, and is Free Open Source). This is relatively easy to get started with. http://graphite.ecs.soton.ac.uk/
This is all very new so we'll be updating things as we learn good and bad techniques.
