lec5 - CS 6093 Lecture 5 Basic Information Extraction Cong...

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
CS 6093 Lecture 5 Basic Information Extraction Cong Yu
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Groups and Projects Primarily supervised by Fernando – P03: Detecting Trends in Facebook/Twitter Feeds • Maggie, Nitin, Quentin – P12: Learning to Rank with Label Noise • Josh Primarily supervised by Cong – P10: Mining Patterns from Status Updates • Devansh, Prayag, Ting – P11: Recommendations in Social Networks • Konstantinos, Rata, Yigit Advice: do the following as soon as possible – Survey the literature – Settle on a well defined problem to attack – Identify appropriate data set to use for experiments
Background image of page 2
Midterm Report 30% of the overall grade Due: 5p ET March 21 st – Deadline is strictly enforced – Late submission policy • One day delay (5p ET March 22 nd ): 5% reduction of score • Two day delay (5p ET March 23 rd ): 20% reduction of score • Three day delay (5p ET March 24 th ): 50% reduction of score • No longer accepted after 5p ET March 24 th . How: email your report to both instructors Format: – 6 pages maximum, including figures, citations, and appendixes – 12pt Time New Roman, single column – Word or PDF only
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Writing Guideline (Suggested) Abstract – A clear and concise description of the project Introduction and Motivation – Describe importance and applications of the problem – Provide motivating examples (real examples are much better) Related Work – Describe existing work in the field • Both research papers and industrial systems – Summarize the solutions of important references Formal Problem Definition and Solution Overview – Formally define the problem being solved – Provide an overview of proposed solution
Background image of page 4
Writing Guideline (Suggested) System Architecture – Provide a system architecture – Describe components already implemented – Describe components to be implemented – Identify off-the-shelf components in the system Preliminary Study – Datasets downloaded and insights obtained from the dataset – APIs examined and experience working with the APIs (i.e., what’s doable and what’s not) – If some preliminary experiments are performed, describe them Bibliography – List of citations
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Any Questions?
Background image of page 6
Today’s Outline Overview of Information Extraction – What, why, and history – IE pipeline & tasks Basic Information Extraction Techniques – Named entity recognition – Wrapper technology Scaling up Information Extraction – Pattern based Iterative extraction – Cost based extraction plan optimization
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
What is Information Extraction? structured record extraction concept recognition
Background image of page 8
What is Information Extraction ID Person Position Place 1 Torsten Suel Professor NYU-Poly 2 Jan Pedersen Chief Scientist Bing Search Concept Recognition – Detecting the occurrences of certain concepts (people, organization, date, etc.) within a document Structure Record Extraction
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Why Information Extraction?
Background image of page 10
Image of page 11
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 80

lec5 - CS 6093 Lecture 5 Basic Information Extraction Cong...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online