Andreas Weigend
Stanford University
Data Mining and Electronic Business
Stat 252 and MS&E 238
Spring 2008

Audio (as mp3): http://www.weigend.com/files/teaching/stanford/recordings/WeigendStanford2008Class9

Sameer Yami - I am interested in working on this page [syami@stanford.edu]"
Aparna Seetharaman - I need to work on this page [aparnaseetharaman@yahoo.com]

Class 9/ Fri 06Jun

Friday 12:15 - 15:15 during finals

Part 1: 23andMe

1215 Alex Wong, Director of Products, 23andMe, Inc., will tell us about the exciting ideas underlying 23andMe, and point to some of the business implications of mining personal genetic data. Following a brief introduction of the company and product, he will discuss genotype data storage and management, social networking with genetic data, and novel online methods of conducting genetic research. He will be joined by lead engineer Alex Khomenko and Founding R&D Architect Serge Saxonov. Alex Khomenko and Serge are Stanford alums while Alex Wong is a graduate of the Harvard CS department.

  • 23andMe brings human genome information to the layman where they can understand their own genome and see various information about possible diseases and also understand their lineage back to thousands of years.
  • Details about human genome project and related topics can be found at :
http://www.genome.gov/10001772
http://en.wikipedia.org/wiki/Single_nucleotide_polymorphism
http://www.thinkgene.com/
http://genome.wellcome.ac.uk/
  • Social networking is built into 23andMe system where people can network with close family members and friends. Social networking combined with genetics can probably lead to eugenics.(http://en.wikipedia.org/wiki/Eugenics)
*

Part 2: Best of Homework 6

1300 Each group has up to 3 minutes to present the one single best thing they did for their HW6.
UI expert Dan Olsen will join Andreas Weigend to judge each contribution (10 points maximum if 3 minutes or less, beyond that maximum shrinks by 1 point every 20 seconds):

Projects


Part 3: Case study: Hitchoo


1400 Yinghan Hu (founder) and Amerson Lin from Hitchoo.com, who flew in specially from Singapore, will conduct a case study and group discussion for us on challenges they faced. We will work through a series of questions, including:

1. Is this model of matching people workable, and what are the main challenges faced by Hitchoo?
2. Should the future strategy be whitelabeled solution, facebook application, or is there a third way?
3. The disincentive (incentive?) of having a business model based on payment credit instead of advertising.
4. Customer acquisition and retaining users.

Familiarize yourself with www.hitchoo.com, and come prepared to share your views/ feedback on question 1.

Part 4: What's next?

1500 The last 15 minutes will be your final opportunity to present any idea you want others to rally behind, getting some of the truly amazing talent of this year's class.
Please sign up before noon on Friday!!

Mingyeow Ng: Discoverio


In this 5 minute session, we will discuss the key problems that the current crop of social networks and platforms have not been able to solve, and what are the limitations that are the bottleneck.

We will then discuss briefly the unique capabilities that Discoverio had been developing in order to help users discover and create meaningful connections. These includes finding new collaborators for projects, business development partners, hiring talent, and forming groups of niche interests.

If you are interested in the discovery space, please check out Discovery is the New Cocaine, a methodology developed to provide systematic way of understanding what Discovery really means to us in the web2 space.

Eric Sun: FaceStat


I'll be talking about data I got from facestat.com (a next-generation picture-rating site) and taking suggestions on interesting ways to mine it.
FaceStat is based on:
  • Market research for individual
  • Allows facial tagging for - relationship status, political affiliation, sexual orientation, trusteworthiness etc.

Dan Olsen: YourVersion

YourVersion is building the world’s best discovery engine to give customers the best way to discover and share personalized, relevant web content related to their interests.