Workshop A: Web-wide Indexing/
Semantic Header or Cover Page

Workshop A:

Web-wide Indexing
Semantic Header or Cover Page

Summary

What to index?

• Use the anchor term used for the HTML links,
• Use the title and headings of the HTML page,
• Use the full text to create an index,
• Use the filename of the HTML resource,
• Word occurrence - URL pairs,
• Inverted indices of keywords,
• Indexes of interesting keywords.

How are indices created?

• Use a robot to scan the Web for new and changed HTML resource,
• Server side support based systems,
• Different frequency of updates.

What is indexed?

• Approximately twenty search engines with accompanying services for parts of the WWW,
• Each covers a part of the Web,
• Gateways to other indexing services such as WAIS.

What is needed?

• How does a naive user start a search? Danger of hierarchical index!
• Hierarchical searching,
• Share information and avoid replication,
• Parallel, fault-tolerant, and scalable index server,
• Language(Natural) independent,
• Find all the _____ having concept/property ------,
• Index other form of resources: images, graphics or sound,
• Capture structure of the Web, hyper-media,
• Common interface,
• Index generation with revision control,
• Establish a Web Indexers' Working Group.