- Hide menu
Just finished listening to these awesome Ischool presentations: NUJJ
I. -What is a location -based service for mobile device users?
Chart of different types of interactions
Face to face communication Flyers, fridge magnets
Graffiti, historical plaques
II. Local questions and answers
You want to eliminate time sensitive questions
III. Travel Planner
-There are a lot of things you might do to prepare for a trip. It doesn’t really make sense to leave those notes in backpack. What if you could leave yourself a note at the entrance to the louvered, a personal note you wrote for yourself, and a travelogue that is location sensitive? It will be delivered right to your mobile service as you arrive.
IV. Recommendation service
-You can leave notes to people who care about it. You get it at the spot where it helps.
V. Lost & found
1. Allows user to tie locations to their notes and allows for location patterns
2. Users may only retrieve public notes from their phones.
NUJJ refers to a service they describe which is platform agnostic.
They are using the iphone to demo.
One of the main problems of the application is scarcity.
When a user sends out a message it goes out through the twitter stream. NUJJ has a special database that stores all event date. It sends out a query for all nudge data.
Better question classification for local Q&A
Is this geo-located twitter? Is this all within network? Can my parents see what I’m doing without being a member?
Scarcity and having an ability to filter.
Requiring it to be social to begin with will be flawed because of the scarcity issues. Delicious works because it benefits me first and there’s a byproduct of helping others. You can’t be all things to all people. Do you find yourself go towards a certain direction?
Buy Music, Make Money
I want a song
But the song’s under lock and key
The pirate bay…
Might feel guild, might need a lawyer
It’s a no brainer
Copy music for friends and earn social incentives, for free.
Existing approaches to fix theses incompatible incentives:
So-called affiliate models. Think Tupperware.
Good: Buyer has a steak in how well a song does. It’s nagging.
Where else can we see economic incentives out in the wild?
Social incentives to be ahead of the curve: Be cool.
Financial incentives: Make a killing on Wall Street.
It’s the same curve as the sale of music.
They wanted to buy a store that…
rewards fan for buying the music they love: an incentive to buy legally.
Appeals to trendsspotters and trendsetters. Give fame in addition to fortune.
Gives buyers a sense of urgency
And encourage buyers to be promoters.
They designed a store where…
1. a song is sold.
2. part of the revenue comes from promotion of that song
Allocating the dollar
50 cent to the artist/music label
50 cent goes to all previous buyers
we want to give early buyers a cut.
You can make sure that once you buy a song you can keep that cut size. Your income growth is linear.
A new song just came out. Say you’re the 5th person to buy it, so you get the 5th largest slice, which is below a 20th of a cent. It will be 50 cents after 1,000 sales, $5 after 10,000 sales; $50 after 100,000 sales.
Make your dollar back after 235,000 copies sold. etc.
How does this compare to a baseline model?
Baseline is Itunes, Amazon etc.
Their model has a network effect.
More value for buyers wit more buyers.
There is a social effect.
Tremendous potential value over baseline for good trendspotters.
There’s fame and fortune for someone who’s good at this.
What we built:
Designed to address the usability of such a model?
Can we make this experience compelling and very simple to use.
Is this model economically feasible? Is the allocation model adjustable to create incentives. What’s the optimal solution.
We did this so can easily test things.
Separation of data, business logic, presentation
Data model – swappable SqLite “stores”
Logic – cut allocation engine
we can swap the allocator and the allocation function.
if we don’t want to allocate things for music anymore you can allocate slices for this engine and it still works for major changes.
Economic feasibility and simulation framework
-works on an actual store instance
performance issues with Django implementation
Useful framework during next steps
we conducted competitive analysis
Do they get it?
Does it work well?
A store should be fun to explore
on left, column of new songs
middle section, songs selling well
right column has social layer. they show the people who perform best and give them some visibility on the site.
Page for one song
description of song, amount, if you buy this song now and performance on the site.
There is no way to compare how a song is doing on the site
they experimented with showing the trend of a song. They did a performance indicator using line graphs to make sense of the numbers. When you’re dealing with the small numbers people have a difficult time imaging what it means and what it means over time.
Break-even point. Buy now and get 0.0475 cents per future copy sold.
Ex. bipod. They didn’t market it as a four-inch box, they said 10,000 songs in your pocket.
They think they found a way to align the incentives they talked about
A mode that benefits all stakeholders
A marketplace where new music is discovered and takes off.
It encourages legal buying and discourages illegal music sharing and piracy.
The project started as a page on a notebook.
they received some feed investment and will make it a startup and launch it as product.
and incentives for news
they are incentivizing jumping on board and taking a risk on music in this case
it’s a crowd-sourcing model monetizing music
why not news?
Question and answer
Music has people can set trends and they are interested in having the economic model flushing out these people with disproportionate influence.
One application that supports multiple points of entry.
Doctor’s information management
Hospitals are information rich environment. They are only focused on note capture and retrieval
They built a mySql database
Request/response to MD:Notes (PHP), my SQL database (emulates EMR)
patient modeled in XML
Schema to map force HL7, mapping transform
It’s driven by the schedule when someone logins.
They have different inpatient settings and the services that are available in the hospital. They are ordered in level of acuity.
once you get the patient level is when you start presenting the patient as xml
One of the reasons we’re using xml is that we need to be able to have the present not just live in our system but interact with automatic hospital system
A healthcare messaging standard for system interoperability
Must be ale to send progress notes fro the EMRI
extending the patient model
supporting model devices
It’s a web application
Why didn’t you look to some kind of mobile application to begin with. They built a voice-prototype and mobile application.
The possibility of a threaded conversation between doctors. Will there be an ability to comment or thread.
Get every new post delivered to your Inbox
Join other followers