Application	DataSet	Example Problem	Conventional Solution	The topicTechnology Solution	Typical Customers
Objective Indexing/ Metadata Enhancement	80,000 articles from the Pennsylvania Gazette	How can I create a high-quality index for this collection?	Manually create an index	Automatically constructs a high-quality objective index	Publishers, content-providers
Topic-based Search	500,000 emails from Enron	What is in these emails? What should I be looking for?	Trial-and-error keyword search	Automatically extracts a topic structure that provides a "map" for efficient search	Law enforcement, Legal discovery, Tech support
Summarization and Visualization	50,000 funded grant proposals from the National Institutes of Health	How can I summarize what we are funding? What are the trends in what we are funding?	Summarize funding amounts using organizational program boundaries	Produces a visual display of funding according to topics, based on the content of the documents, allowing interactive "drill-down" querying by topic, program, year	Federal funding agencies
Organization/Expert Assessment	1 million technical articles from MEDLINE or CiteSeer	What research areas is this university/ company/ research organization/ country strong in?	Manually investigate each university, company etc. at a time - use predefined categories to categorize their research	Automatically builds a database of expertise organized by individual or organization, supporting interactive flexible querying at any level of aggregation	Corporate strategic planning, investment firms
Taxonomy Extraction	10 million Web pages related to a particular topic	How can I automatically create a taxonomy to help users navigate this collection?	Manually construct a taxonomy and manually classify each Web page into a category	Automatically constructs the taxonomy and automatically assigns Web pages to categories in the taxonomy	Web portals, Search engine companies
Hot Trend Detection	200,000 Web blogs	What are the new emerging trends?	Hire editors to read the Web blogs daily and spot new ideas	Automatically tracks and categorizes Web blog discussions and produces a "report stream" on new topics detected	Market research firms, online content providers
Topic Translation	100,000 documents in a foreign language	What are these documents about?	Employ a translator to search the documents	Topics and relevant documents automatically detected, only relevant documents sent to translator	Intelligence agencies, Web portals