|Hosts: Leo Laporte and Amber MacArthur|
Guest: Mr. Tweet, aka Ming Yeow Ng
Topic: Mr. Tweet
Published: December 25th, 2008
Net@night 80 - Mr. Tweet
- Mr. Tweet, aka Ming Yeow Ng
- blog.mrtweet.net for greater detail on the mechanics of Mr Tweet; algorithms etc..
- In the early 90s people were stil takling about routers and switches whereas now the medium has matured they're talking about community and information, debate. Leo says a similar thing happened with stereophiles and audio systems. The conversation progressed from Total Watts RMS etc. but about their favourite bands and such.
Interview with Mr. Tweet
- Mr Tweet is your personal networking assistant, it finds out who you should follow on Twitter.
- Twitter is a simple system, it's users invented conventions such as the @ reply for which Twitter then added support.
- Finding people to follow on Twitter is not easy and Twitter doesn't help by proving functionality
- No one Twitterer can engineer a feature, it needs to be a grassroots thing (like the # hashtags)
- Mr Tweet is currently two people
- There is a noise problem on Twitter, browsing people on Twitter is very difficult, Mr Tweet was created as a solution to these problems.
- To get help from you follow Mr Tweet on Twitter and within a few days/a week you will be contacted and asked what you want Mr Tweet to do for you
- You can find new people to follow
- You can find people within your network that you might want to follow back.
- Figuring out who you should follow is a two part process.Data collection from Twitter "If it's feeling well then we get the data". The computation of that data
- Twitter development team are incredible responsive and helpful as they realise the 3rd party ecosystem is essential to their success in a way it isn't for Facebook.One killer app like Mr Tweet can make your service grow
- Mr Tweet has created nearly 400,00 new connections.
- Amber is pleased with her suggestions and finds that many of the suggested people were people she thought she was following already
- The suggested people have statistics beside their names e.g. how many followers frequency/amount of updates
- There are many different kinds of Twitter users, using it in different ways
- Tim O'Reilly follows about 100 people, does a lot of re-tweeting, rewarding the people that give him good information by spreading it
- Loic LeMur follows "like 15,000 people", if you follow both Tim and Loic, Tim's followers will carry more weight because Loic follows probably a lot of spammers and bots. Mr Tweet is trying to ascertain credibility.
- Ratio of followers/followed doesn't necessarily mean that much: people who follow many people may be earer to learn and get information, or they could be spammers,
- "We try to see who the user thinks is a credible person, not so much who we think is a credible person, if we were to impose that measure... everybody would see Tim O'Reilly " (27:05)
- "My one year old kid has no influence on the rest of the world, but he has tremendous influence on me" (28:08)
- Amazon uses user generated data in their recommendations, how like user A user B is. how to bring these two similar people together
- slideshow on the ideas behind Mr Tweet
- Andreas Weigend a Mr Tweet advisor and former chief scientist at Amazon.com
- Company site of the people behind Mr Tweet (just an info page at the time of transmission)
Picks of the Week
Website of the Week
Video of the Week
Podcast of the Week