Topikality is startup based in Sydney, Australia. It has an online application that finds articles on topics that interest you and filters out the useless stuff. Unlike Google news alerts which search for articles based on keywords alone, Topikality allows you to vote articles up or down, and learns from your votes by presenting you more filtered and relevant results. Topikality’s statistical machine learning technology analyses approximately 200 of the article’s features, and draws comparisons against previous articles that you liked (and didn’t like) – looking for trends.
Topikality is in public beta. You sign up for an account and start creating your “topiks” (see below).
In the next screenshot, you see the results of the topik I created and called “Technology”. I asked Topikality to look for articles with the word “startup”.
To learn more about Topikality, go to their demo posted on YouTube.
Here is the Q&A with Philip Scott, co-founder of Topikality.
When did you start Topikality?
In December 2007, my co-founder, Richard Heycock and I had our first discussion on how to make sense of all the information on the Internet. It seemed to be a growing problem for everyone and there wasn’t any available solutions. Richard had ideas on developing the smarts to learn what was relevant to the user and eliminate the clutter. We thought if what we create is real and scalable then it would be of valuable to people. Richard started work immediately on the prototype. So technically we started work in December 2007 and we have been working on the project ever since.
What was your inspiration (apart from solving the problem of information overload)?
Richard was initially inspired by a conversation with his partner about researching information relating to her work as a nutritionist and blogger. Given that Richard’s entire career had involved massive datasets and some machine learning, it seemed like a good idea to combine these two things.
My inspiration was trying to keep up-to-date on the things that really interested me. The traditional media was too general, industry publications provided a single view of the world, and online systems driven by popularity weren’t what I wanted. Searching online everyday was too time consuming.
The concept was quite appealing to me. A system that knew what I liked, continually researched the Internet for me and emailed me a set of relevant articles each day. We set out to try to make this a reality.
We thought, even if nobody else likes it then at least it would be useful for ourselves.
What is your background?
I was co-founder of the Prosper CRM software business that specialised in CRM with real-time integration to the banking platform of Credit Unions and Community Banks. It was marketed in the US by Harland Financial Solutions as Touche Sales and Service. Our biggest competitors were Onyx and Pivotal. I left the business in 2006.
How many people work at Topikality?
Two: Richard and I. Along the way we’ve had lots of help from friends and family.
Where do you want to see Topikality by the end of 2010?
Out of Public Beta, much higher user numbers and integration to other web applications.
Since you are both frustrated cyclists (I follow professional cycling myself but I don’t have Lance Armstrong legs), is this what you do during weekends? Go on long rides? Have you ever done the “amateur” Tour of Flanders?
Being professional cyclists is only a dream. Our ‘About us’ page says we are frustrated professional cyclists but its only said in jest. Richard raced when he was younger. I played football (i.e. soccer). Now Richard gets around town on his bike and I am a social rider. That is, I ride regularly but only when there is a cafe at the end with good coffee!