Thought Flow

Tag: data mining

  • Becoming a data scientist

    Data vs. Spock (70/365)
    Data vs. Spock (70/365) by JD Hancock (CC-BY)

    It is is interesting how a job title carries certain connotations and how it can even be tailored to different situations. Sometimes I have called myself a Consultant, because this is what consultancies expect. Sometimes I have called myself a Developer, because that is what hiring companies expect. Sometimes I have called myself a Software Engineer, because it sounds good to Google and Amazon. All of the titles are technically correct (I hope), but create slightly different associations and expectations.

    In my current position at Receiptful, things are a bit different. I am now a Data Scientist. This is not a title I applied for or have ever used about myself. If someone called me a data scientist before I started at Receiptful, my response would be: “Well, I don’t feel like a data scientist. I’m a developer with knowledge of data mining.”

    This is the first time I have not made up my own title. It was handed to me and now I have to live up to it. The funny thing for me personally is that “data scientist” is actually a title I have aspired to have for a long time, but I just never felt that I was entitled to call myself a data scientist. Data science is such a hot topic right now, and I really do not want to seem fake.

    So for now, I will continue to build cool stuff and hopefully, the title becomes self-evident after a while. Time will tell :-)

  • Antecons

    “Release fast and iterate“. In some parts of the computing world this is the mantra and I think it’s time for an early announcement: Metacircle has now released an actual product (kind of). It is called Antecons, short for antecedent – consequent. Antecons is a small and simple component that suggests items based on previous usage data (such as purchases) and it is currently being tested on vildmedvin.dk.

    Why did I create Antecons? Well, there are several reasons, actually.

    • I love data mining and business intelligence
    • Good recommendation engines improve conversion rates and thus profit for e-businesses.
    • Recommendation engines are not available to everyone at a low price. Most current solutions like avail.net target big companies and corporations. Antecons targets everyone.
    • People have already shown interest in Antecons. This includes my faithful partners at MCB as well as the founder of Remove The Background

    Let’s keep it real. The business intelligence market is very competitive and many companies are years of research ahead of me. I do not plan to steal their customers. I plan on making advanced technology available to the masses.

    In its current form, Antecons is simply a .NET DLL component with a public API for interacting with Antecons. It needs access to an SQL server and will thus only work with ASP.NET applications. I am currently exploring the possibilities of making Antecons a software-as-a-service (SaaS) product and a proof-of-concept prototype is being built for the Google App Engine, although it has not been finally decided where to host the first SaaS version. You can check it out here.

    If Antecons gains traction, I will give up consulting and work solely on making it an excellent and easy-to-use product. For now, it has been a major personal success to see something get out there. I hope the interest continues.