작성자: 과거 관리자
작성 날짜: 2012/02/23 (목) 오후 5:49
정보문화학에서 아래 강좌를 개설하여 홍보를 요청하여 안내해 드립니다.
수업이름은 창의연구실습 (금 1,2,3) 이고, 현재는 행정적인 이유로 김은미 교수님이 담당으로 되어 있을겁니다.
조만간 Shilman 으로 바꿀 예정입니다.
Lectures
Class sessions are three hours in length, and most will contain one or two 30-minute lectures on a specific topic in Collective Intelligence, as well as interactive workshops, discussions, and hands-on lab work. Guest speakers will also help keep things fresh.
Week-by-week topics. Check back for slides as the class progresses.
• Wk 1: Introduction
Who am I? Who are you? What is Collective Intelligence? What are compelling examples? Why is it useful? Why is it a great thing to study? What can you expect to get out of the class? How will the class work?
• Wk 2: Survey Methods / Data Mining
Surveys are a very direct and mature way to understand a population. What are the elements of a survey? What are the principles of survey design and analysis? What is data mining and how do you do it?
• Wk 3: Visualization and Analysis
What are the principles of visualization design? How to build interactive, animated visualizations?
• Wk 4: Social Mechanics
How do social mechanics encourage and harness constructive crowd behavior? Open source? Wikipedia? Quora? Game mechanics? What principles can we take away from these systems? What are some problems that could benefit from these kinds of mechanics?
• Wk 5: Creativity and Brainstorming
How does creativity manifest itself in a crowd? How do we conduct a productive brainstorm?
• Wk 6: Project Management
How to manage a software project? What are waterfall, iterative design, agile development, and lean process? How do you incorporate crowd participation in your process? How does this apply to your project?
• Wk 7: Prediction Markets
How to use crowds of people to improve predictions? What are some successes and failures of this approach? When is it useful and what are rules of thumb for designing it?
• Wk 8: Text Mining
What are effective techniques to spot entities, topics, and sentiment in text? What is machine learning? What can these tools do today?
• Wk 9: Analytics and Community Metrics
What are useful metrics for understanding aggregate behavior on your site? How do you measure these and interpret the results? What are useful community metrics?
• Wk 10: Project Update
Status presentation, initial findings, peer/panel feedback.
• Wk 11: Predictive Analytics / Collaborative Filtering
How do we model human preferences to build personalization systems?
• Wk 12: Reputation Systems
How is social data used to build reputation systems? How do PageRank, EdgeRank, Klout, PeerIndex, StackExchange work? What is the future of reputation systems?
• Wk 13: Crowdsourced Labor
How do labor markets like Amazon Mechanical Turk, oDesk, etc. work in practice? What are the challenges? What are techniques for ensuring high quality results?
• Wk 14: Meme-Tracking and Viral Marketing
How do ideas spread through communities? What are the characteristics of viral successes? What are effective methods for spreading ideas? How do you track them?
• Wk 15: Project Presentations
Final presentation, findings, peer/panel feedback.
• Wk 16: Reflection
Overview of what we learned, overview of Collective Intelligence topics that we didn’t cover. Gather feedback for future iterations of the course. Discuss follow-on work and future directions.
=== ============================
Instructor
Michael Shilman is an Adjunct Professor in Seoul National University’s department of Information and Culture Technology Studies and Founder of Lab80, a technology workshop in Seoul’s Hongdae neighborhood.
Michael was founding member and Chief Scientist of Wize, a Mayfield and Bessemer-funded Silicon Valley startup. Wize survived the financial crisis of 2008, grew to profitability, and was successfully acquired by the world’s largest price comparison shopping engine Nextag in July 2010. At Microsoft Research, his handwriting, document, and image recognition technology shipped in Windows, Office, and Bing, and led to 20 papers, 18 patents, and demonstrations by Bill Gates on CNN and BBC. He prototyped and engineered a scanning-and-speech interface for commercial check processing which currently processes over $5B every day.
Michael was a Mayfield Fellow and Hertz Fellow at UC Berkeley where he received his BS, MS, and PhD in Electrical Engineering and Computer Science.