SE 305: Topics in Web-scale Knowledge Harvesting
Course Flyer: https://sites.google.com/site/se3052014/SE305_Flyer.pdf
Instructor: Partha Pratim Talukdar [ppt@serc, http://talukdar.net]
Time: TTh 2pm – 3:30pm [first class: Aug 5, 2014 (Tue)]
Location: SERC 202
Number: SE 305
Term: Aug – Dec 2014
This is a research-focused course where we shall read papers to understand and extend state-of-the-art statistical machine learning techniques for various problems in web-scale knowledge base (KB) construction, e.g., entity categorization and resolution, relation extraction, distant supervision, KB inference, curriculum learning, learning from limited supervision in KB, representation learning for knowledge harvesting, etc.
Particular emphasis will be given to techniques that can scale to large datasets, preferably using distributed frameworks such as Hadoop.
Students will be evaluated based on their presentation and participation in class, and also a synthesis report or project (as preferred by the student) to be submitted at the end of the semester. Given the interdisciplinary nature of the topic, interested students from all backgrounds are welcome to participate.
Interest in any of the following: machine learning, natural language processing, large-data processing. Necessary background will be provided.