Optional Extra Credit Assignment 1: Metadata
(Up to 40 extra points may be awarded for successful completion of this assignment)

Readings: Holland et al., WordNet chapters, Bates '88, Svenonius, Pitts-Moultis and Kirk, OI chapters 1-8.

The goal of this assignment is to give you hands-on experience with three aspects of metadata:

1. Metadata coding (using XML and Dublin Core)
2. Organizing information via metadata, distinguishing hierarchical versus faceted organizations
3. Lexical relations and their relationship to conceptual relations, via WordNet.

(1) XML Data DTD Construction and Dublin Core Description

This assignment is intended to help introduce you to XML DTD design and also to help you think about and carry out descriptions of network and other information resources using the Dublin Core metadata framework. Information on the Dublin Core descriptive elements and their use can be found at http://purl.oclc.org/metadata/dublin_core/. Discussion of XML DTDs and their elements can be found in the reader, and also via the Robin Cover SGML/XML web site. Note that not all items will require all descriptive elements.

The assignment has three parts:

• (A) Create an XML DTD that has all 15 elements of the Dublin Core as XML elements.
• (B) Create Dublin Core descriptions for each of the items listed below.
• (C) Format your answers as well-formed XML documents using your DTD.

 

 

The items that you are to describe are:

• The textbook for this course.

Create one Dublin Core metadata description of Baeza-Yates & Ribeiro-Neto, Modern Information Retrieval.
• The UC Museum of Paleontology Website: http://www.ucmp.berkeley.edu/

Create one Dublin Core metadata description for each of:
• the main web page
• the main online exhibits page
• the Phylogeny sub-site linked to from the main exhibits page.
To create the metadata description simply use a word processor and mark up your description of these items using the DTD created in part 1. Your description should go into the appropriate elements defined by your DTD -- the content should be based on what you find at the OCLC site You shouldn't need to read all of the information on the web sites to be able to do these descriptions, just enough so that you have a good idea of what they are about and how to describe them.

Be sure to include your DTD along with your marked-up descriptions.

(2) Hierarchical Versus Faceted Classification

Say you are making a classification system to describe recipes. Assume your classification has to include the following dishes:
1. Broccoli and Potato Soup
2. Chicken Curry
3. Grilled Salmon
4. Pad Thai (spicy noodles, shrimp, and vegetables)
5. Paella (spanish rice, sausage, and chicken)
6. Chocolate Cheesecake
7. Buttered Baby Carrots
8. Strawberry Milkshake
Assume also that following types of metadata have been assigned to each recipe:
• Main Ingredient (e.g., Pasta, Poultry, Cheese, Chocolate, Seafood ...)
• Cuisine (e.g., African, American, Asian ...)
• Preparation Method (e.g., Bake, Broil, Quick, Steam ...)
• Season/Occasion (e.g., Autumn, Picnic, Thanksgiving ...)
• Course/Dish (e.g., Appetizers, Dessert, Soup, Snack ...)
Also, some of the categories in the Main Ingredient facet have hierarchical structure. For example, under Seafood are the subcategories Fish and Shellfish, and under Poultry are the subcategories Chicken, Turkey, and Duck. (Feel free to add more hierarchical structure to other categories as necessary.)

(A) Create a faceted classification system that can be used to describe these dishes. Show how each dish can be classified within the classification system. Your design must take into account the fact that at least one facet is hierarchical (Main Ingredient). You don't have to show the entire classification system; only those parts that are needed to classify the dishes shown. Briefly explain/justify your classification system.

(B) Create a hierarchical classification system that can be used to describe these dishes, and show how the dishes can be classified within the system, using all of the types of metadata shown. You don't have to show the entire hierarchy; only those parts that are needed to classify the dishes shown. Briefly explain/justify your classification system.
 
 

(3) Lexical Relations and Conceptual Relations using WordNet

Use the WordNet HTML forms interface (or any of the other interfaces available to explore a bit of WordNet.

(A) Select a noun that has at least two synsets assigned to it. Draw a sketch of links to/from this synset to other synsets. For at least two senses of the selected word, show at least one hypernym, one hyponym, and one other type of lexical relation.

(B) Name a conceptual relation that can associate a concept with the word you chose but which is not a valid WordNet lexical relation. For example, for the word tennis, a conceptual relation that might hold between it and racquet is instrument. (Do not use instrument as your example!)

---