Analytics in Khan Academy Talk on TED.com

A TED.com talk by Salman Khan has been getting a lot of publicity recently:

http://www.ted.com/talks/salman_khan_let_s_use_video_to_reinvent_education.html

The talk is about the Khan Academy (www.khanacademy.org) which is a relatively new online education site. While some of the coverage on changing education and "flipping the classroom" might be a bit over the top (is asking students to watch a video before they come into class really that different to asking them to read the next chapter of their text book?) the data analytics behind the system looks really interesting. You need to get about half way through the video to see it but there are really nice visualisations showing student performance etc (I thought the graph showing the problems with streaming classes was a really powerful example of visualisation).

Gmail Priority Inbox & Explanation

Gmail Priority Inbox was launched a little while ago and is a nice system that attempts to learn to recognise the emails that are important to you and flag these as such - sort of reverse spam filtering.

The system uses a simple linear regression model to predict "importance" but customises the model, and the threshold used to get emails over the line into a user's priority inbox, based on a user's interaction with their mailbox. There's a nice paper about this by Google employees here.

The other nice thing I have noticed about this recently is that Google have included some nice explanation features - so when you look at an email you are told why it was classified as important. An example is shown below.

This is a nice example of an attempt to allow users look "inside" a machine learning-based application. It does, however, raise an interesting question. Given that users can now access information as to why particular classifications are made, what can they do about it? In the case of Google Priority Inbox, at the moment nothing. While we can tell the system that emails are important or not we cannot directly influence the decision making. This illustrates clearly the problem of balancing insight into the behaviour of machine learning-based applications with the resulting frustration users feel from not being able to change that behaviour.

Drop-in Session

Hi folks,

I hope you have settled into the new semester well and are enjoying your new modules.

To give students a chance to ask any questions or address any issues that have arisen for them on the programme I will be in room A3-020 this Thursday (10th March) from 17:45 to 18:30. There is no need to attend, but I will do my best to answer any questions or address any issues that arise.

We will have another similar session towards the end of the semester.

Regards,

Brian.

No such thing as bad results - just creative stats

For light relieft this week, statistical proof of the unstoppable rise of Brazilian rugby (subtitles).

http://www.youtube.com/watch?v=AKnQn-8PFII

Three Interesting Data Analytics (and Visualisation) Things

Three interesting analytics finds online this week folks.

First, an article from the Guardian newspaper about the new publicly available "crime map" from the UK (www.police.uk):

"All the crime map shows up is Whitehall's pointless zest for data" by Simon Jenkins

It seems to me that the author misses the power of making data accessible and the real problem is that the data being made available here is overly curated.

The next two are both visualisation tools. Hipmunk (www.hipmunk.com) is a flight search tool that visualises flight availability in an interesting new way. Also their derived variable "agony" would be well worth understanding more (how can you define such a subjective measure?). Prof. Barry Smyth from UCD has an interesting review on his blog.

Finally (and thanks to Colman McMahon for pointing this one out), LinkedIn have launched a tool called LinkedIn Maps (inmaps.linkedinlabs.com) to help you visualise your professional network. Again it would be interesting to know how they choose node colours, sizes and positions relative to each other. A very interesting tool though.

Data + Database Design for Data Analytics

Dr. Pierpaolo Dondio should have been in touch by email re this module.
Classes will start for this module on Wednesday Feb 16th
All material will be covered as expected

Timetable Update for Semester 2

Apologies for the lateness of the change - this is due to
lecturer timetable conflicts on another programme

The timetable for S2 will operate as follows:


Tue 18:30 20:30 KA - 3-020 Enterprise Systems Integration (Ronan Bradley)
Wed 18:30 21:30 KA - 3-020 Data and Database Design for Data Analytics (Pierpaolo Dondio)
Thu 18:30 21:30 KA - 3-020 SPEC 9270 Machine Learning (Brian MacNamee)