Project meeting at Stanford
Date: 02/11/99
Attendees: Rajeev Motwani, Svetlozar Nestorov, Sebastien Brion, 
Dick Tsur, Yue Zhuge

We discussed more extensively about XML and our graph model to represent
XML documents.  In particular, we discussed the following two topics:

1. Links in XML. We found out that there are many types of links XML
allows to define, they may be very useful and we plan to study them
further after the meeting. We will also find out the differences between
XML links versus HTML links.

2. We decided that our current graph model for XML contains the 
following items:

* Two type of nodes
	- Element node: each element node contains the tag of 
	the element, its attributes, and a list of links to its sub-elements
	- Text node: each text node contains free text (PCDATA) as its value
*Three types of links
	- Sub-element link
	- IDREF link
	- Remote link (across documents)
Links may need to have labels on them.

After we study in more detail about XML links, we may decide to refine 
the model.

In this meeting, we also studied resume samples Yue prepared. Those
resumes are all downloaded from the web and they belong to three different
groups of people: EE/CS engineers, lawyers and actors/actresses.  Lists of
key words used in the three groups of resumes can be found here. We think
these are good sample data and we may start to = build simple testing
system using them. (For example, map them into XML documents, come up with
some DTD for them, etc.)