Iwsearch

From Inference Web

Overview

goals

• discover PMLP data published on the Web
• validate PMLP data
• build a search engine for PMLP instances on the Web

People

• Li Ding
• Cynthia Chang
• Zhenning Shangguan

Architecture

• discover component
  • crawl a given web directory
  • input: a set of starting URLs; a set of regular expression patterns for allowed result URL; max crawl depth
  • output: list of URLs

• validation component
  • validate a markup page or content in a markup page
  • RESTful webservice
  • input: a URI (an instance or a document)
  • output: validation result (ok, warning, error); detailed explanations (list of warning and error message).

• ping component
  • check the current status of a published URL, keep ping history
  • RESTful webservice
  • detect changes of URL
  • validate (depends on validator web-service)

• index component
  • incrementally update "index" of PML document