Slides

hipshorseheadsΔιακομιστές

17 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

108 εμφανίσεις

Diabeat.us

Diabetes lifestyle management


Andrew
Alles

Victor Benjamin

Robert Erikson

Xiao Liu

Introduction


With the emergence of web 2.0, individuals have
been heavily using the Internet to share ideas and
experiences


In particular, support groups are quite prominent and
useful to Internet users, especially concerning medical
problems


However, much of the content existing on the
Internet today is disjointed


Requires users to visit multiple sources


Poor organization can make content hard to find

Objectives


Diabeat.us seeks to advance the
accessibility
of
diabetes related resources to Internet
users


Diabetes related news and recipes


Educational books and videos


Web community analytics


Content represented must come from a set of
diverse, yet credible information sources


Current leading diabetes information sources (e.g.
diabetes.org)


Various popular web APIs contribute content and form


Information visualized from patient
-
generated web content


What’s so special?


Functionalities


Meta
-
search
across multiple renown diabetes
resources


Nutritional Information, Recipes


News, Research Progress


Videos, Multimedia


Social media analytics


Sentiment analysis on potential diabetes treatments


Social network analysis to identify credible and helpful
individuals in diabetes support communities



Competitor Analysis







Much overlap exists between currently existing web
communities


However, analysis of user generated content is not featured


Such analysis could benefit patients by quickly assessing the
opinions and experiences of entire web communities. It gives us
an edge where competitors cannot “
diabeat
” us.


Website

Basic Diabetes

Information

News

Recipes

Multimedia

Drug
Sentiment
Analysis

Social
Network
Analysis

Diabetes.org









X

X

Diabetes.co.uk









X

X

Dlife.com









X

X

Diabeat.us













Business Models


Advertisement


Google
Adsense


Amazon Referrals


Youtube

Channel Partnership


Subscription
-
based content


Sentiment analysis of treatment discussions


Social network analysis of web communities


Architectural Components


Web Design and Hosting


Amazon EC2


Apache Tomcat


Windows Server 2008


MSSQL 2008


APIs


Twitter


Google
Adsense
, Feed, Search


Google Android


Amazon


Facebook


Youtube


Wikipedia


Flickr


IBM Many Eyes (SNA Visualization)



Analytics and Novelty


We perform text mining of user
-
generated
web content to quickly assess what
people are saying about diabetes


Sentiment analysis on drug and treatments
can quickly let patients know how others
experience and feel
various treatments


Member Contributions


Member Contributions


Andrew
Alles



Server Admin, App development


Victor
Benjamin


Web and backend programming


Robert
Erikson


API research and implementation


Xiao
Liu


Analytics and API programming



All
members assisted each other with various tasks


E
ach member participated in collection and extracting
e
xisting web content


Bug fixes and design changes handled collectively




Future


Future


Advance analytic services and content


Expand website to include Chinese content


Development of analytics for Chinese content


Establish better social media presence