careers

Probabilistic logic in document collections

Documents often come in groups of pages that are highly related to one another. For example, they might follow a page sequence, have a common title, use similar layout and fonts, or include a single name in multiple form fields.

These document relations operate like rules, but with exceptions and differing reliabilities. We'd like an intern to help  analyze document collections through probabilistic logic.

Requirements:

• Senior graduate student
  
• Working at forefront of probabilistic logic -- can help do learning and inference of soft-logical relationships over collections of documents