Collective Intelligence Week 1

Slides:



Advertisements
Similar presentations
Introduction to CS170. CS170 has multiple sections Each section has its own class websites URLs for different sections: Section 000:
Advertisements

Equations in Simple Regression Analysis. The Variance.
Copyright © 2008 by Pearson Education, Inc. Upper Saddle River, New Jersey All rights reserved. John W. Creswell Educational Research: Planning,
N318b Winter 2002 Nursing Statistics Specific statistical tests: Correlation Lecture 10.
Introduction to Regression Analysis. Two Purposes Explanation –Explain (or account for) the variance in a variable (e.g., explain why children’s test.
Business Discipline Breakout Session Summer 2000 ION Conference Facilitated By: Marcy Satterwhite.
CMSC 671 Principles of Artificial Intelligence Course Overview Fall 2015.
Experimental Research Methods in Language Learning Chapter 11 Correlational Analysis.
Welcome to CS 115! Introduction to Programming. Class URL Write this down!
FACULTY ADMINISTRATIVE SCIENCE AND POLICY STUDIES ADS511 RESEARCH METHOD AND DATA NALAYSIS CORRELATION ANALYSIS.
INTRODUCTION TO PROGRAMMING ISMAIL ABUMUHFOUZ | CS 146.
Web Syndication Formats Seminar Week 1 Old Dominion University Department of Computer Science CS 791/891 Spring 2008 Michael L. Nelson 1/16/08.
Web Programming Week 1 Old Dominion University Department of Computer Science CS 418/518 Fall 2007 Michael L. Nelson 8/27/07.
Web-Based Information Retrieval Week 1: Administrivia Old Dominion University Department of Computer Science CS 895 Spring 2013 Michael L. Nelson 01/15/13.
CSE6339 DATA MANAGEMENT AND ANALYSIS FOR COMPUTATIONAL JOURNALISM CSE6339, Spring 2012 Department of Computer Science and Engineering, University of Texas.
Research Methods: 2 M.Sc. Physiotherapy/Podiatry/Pain Correlation and Regression.
“Crowdsourcing” a Textbook: 120 Student Authors Writing on a Wiki
Joan Donohue University of South Carolina
Course Overview - Database Systems
CMSC 471 Introduction to Artificial Intelligence section 1 Course Overview Spring 2017.
Introduction to Information Retrieval Week 1: Administrivia
Web Server Design Assignment #4: Authentication
Course Overview - Database Systems
Introduction to Information Retrieval Week 1: Administrivia
Web Programming Assignment #1: Basic BBS Operations
Andy Wang Operating Systems COP 4610 / CGS 5765
Web Server Design Assignment #5 Extra Credit
Economics of Information Week 1
Introduction to Digital Libraries Assignment #3
Web Programming Assignment #1: Basic BBS Operations due: 9/28/2010
Andy Wang Operating Systems COP 4610 / CGS 5765
Introduction to Digital Libraries Assignment #2
Accelerated Introduction to Computer Science
Introduction to Comparative Effectiveness Course (HAP 823)
Web Programming Assignment #4: Searching, Notification & Upload
Web-Based Information Retrieval Week 1: Administrivia
Introduction to Digital Libraries Assignment #3
Introduction to Digital Libraries Assignment #3
Web Server Design Assignment #2: Conditionals & Persistence
Unit XI: Data Analysis in nursing research
Old Dominion University Department of Computer Science
Web Programming Assignment #2: Expanded BBS Features
Web Programming Assignment #2: Expanded BBS Features
Introduction to Digital Libraries Assignment #2
Introduction to Information Retrieval Assignment #3
Introduction to Digital Libraries Assignment #3
Web Programming Assignment #3: Admin and User Functions
Web Programming Assignment #4: Searching & Notification
Web Programming Assignment 4 - Extra Credit
Technologies of Google Seminar Week 1
Andy Wang Operating Systems COP 4610 / CGS 5765
Web Programming Assignment 4 - Extra Credit
Web Programming Assignment 4 - Extra Credit
Lecture 1a- Introduction
Descriptive statistics Pearson’s correlation
Introduction to Digital Libraries Assignment #2
Web Programming Assignment #1: Basic BBS Operations
Introduction to Digital Libraries Assignment #3
Introduction to Digital Libraries Assignment #3
Web Server Design Assignment #5 Extra Credit
Web Programming Assignment #1: Basic BBS Operations
Introduction to Digital Libraries Assignment #1
See Table and let’s do it in R…
Old Dominion University Department of Computer Science
Web Programming Week 1 Old Dominion University
Introduction to Digital Libraries Assignment #4
Introduction to Digital Libraries Assignment #2
Web-Based Information Retrieval Week 2: Administrivia
Web Programming Assignment #3: Admin and User Functions
SWE 632 User Interface Design and Development
Presentation transcript:

Collective Intelligence Week 1 Old Dominion University Department of Computer Science CS 795/895 Spring 2009 Michael L. Nelson <mln@cs.odu.edu> 1/14/09

Goals We will: learn to program in Python survey mathematical techniques for mining the web and be able to use them in actual applications cover each chapter in the text book note that each chapter is a survey; we could spend a semester or more on every topic…

Administrivia This is a programming class! You should be able to teach yourself a new programming language from the book examples You will work in teams of one or two people mixes (g/u, g/g, u/u) ok assignments are the same regardless of group size

Administrivia 2 Pick teams wisely teams will exist by mutual consent only at any time, teams can split up, but no new teams will be formed after the first assignment is due no team member swaps ex-team members will have access to their shared code base

Administrivia 3 Important URLs Class homepage: http://www.cs.odu.edu/~mln/teaching/cs895-s09/ http://list.odu.edu/listinfo/cs895-s09/ Class homepage: readings are posted demo days are posted

Grading 10 problems, 10 points each Days of in class demo are posted 1 problem each from chapters 2-11 points recvd for each problem at the discretion of the instructor each group must email to the class list which problems they are planning to cover instructor has the right to prohibit exercises deemed “too easy” and to approve exercises not included in book since len(groups) > len(problems), overlap is ok (but variety is encouraged) Days of in class demo are posted demo day 1: chapters 2,3,4 demo day 2: chapters 5,6,7 demo day 3: chapters 8,9,10,11

Development Environment Assignments must be demoed on the machine mln-web.cs.odu.edu same uid/passwd as departmental unix machines; I will have accounts set up shortly Assignments will be written in Python

Extra Credit 10 points available for each person (not group) for: asking a question on the email list answering a question on the email list sharing tips Only “significant” messages will be counted (instructor’s discretion) e.g., “how do I login?” will not count

Why do we care about mining the Web? People in aggregate can be very smart… sometimes referred to as the “wisdom of crowds” http://en.wikipedia.org/wiki/The_Wisdom_of_Crowds …and sometimes not. http://en.wikiquote.org/wiki/List_of_misquotations http://www.youtube.com/watch?v=h6sj89xgnl4 pagerank image from: http://hamletbatista.com/wp-content/uploads/2007/10/pagerank.jpg

Chapter 2: Recommendations

Comparing Reviews of 2 Movies

Comparing 2 Reviewers

2 Highly Correlated Reviewers

Correlation A measure of a linear relationship between two independent variables from: http://en.wikipedia.org/wiki/Correlation

Correlation In the book, we use “Pearson’s Product Moment Correlation Coefficient” http://en.wikipedia.org/wiki/Pearson_product-moment_correlation_coefficient Many, many, many other coefficients possible e.g, for rank data, we often use: Spearman’s Rho Kendall’s Tau http://en.wikipedia.org/wiki/Rank_correlation

Look at data first!!! Pearson’s only works for linear (i.e., normal distributions) non parametric correlation methods needed for non-normal distributions http://en.wikipedia.org/wiki/Correlation#Non-parametric_correlation_coefficients Example: Anscombe’s Quartet http://en.wikipedia.org/wiki/Anscombe%27s_quartet

Recommendations Based on Past Rankings Note not all reviewers have seen all movies!

“If you like X, you probably won’t like Y.”

Item Based Recommendations (i.e., movies instead of critics)

To Do for Next Time… Subscribe to the class email list Submit group info to class list I’ll email to the list when accounts on mln-web.cs.odu.edu are set up