In 2018 I participated in a machine learning hackathon despite having no machine learning or data science experience. What attracted me to it was the Natural Language Processing aspect of machine learning. Anything that has to do with languages and computing (such as computational linguistics) has an irresistible draw for me, even though - and especially because - I've never done it in any of my software development jobs. It is very far from my area of specialization. My only foray into computational linguistics so far has been experimentation with the word2vec Python package. That was true at the time of the hackathon, and it is still true 7 years later. Why? That's a long and not particularly interesting story.
But I had an idea for an application, so I thought, why not try to go and implement it during the hackathon. What do I have to lose?
My second reason was that it was an all-women hackathon, so I didn't have to worry about the judgement of tech bros. I could build an application as embarrassing as I wanted. Turns out, as far as that criteria is concerned, I succeeded in spades.
On Sunday, at the end of the weekend-long event, when the teams presented their results, I had an eye opener: at a machine learning hackathon - as opposed to a regular one, where people build apps - people did NOT build apps. Rather, they wrote Python and R scripts and presented the results of the analysis performed by those scripts. Most of their presentations consisted of slides of the results. Me, on the other hand, I knew how to write software applications, so that's what I did. In that weekend, I built an application with a simple front-end that displayed some "interesting stuff" I found out via my word2vec experimentation.
By "interesting" I mean interesting to me, but apparently not to the members of the jury, nor to anyone else in the room. At this hackathon, unlike any others I've been to, both the participants and the jury were able to assign points to each project, and then the points were announced publicly. (By the way, I don't recommend this practice. It would be enough to announce the first 3 winners, but it does no one any good to announce the points scored by non-prize winners and to each team to know their exact ranking.) To no one's surprise, my project was last. It didn't surprise me either, because, as I said above, I had no experience in machine learning. I was more or less the only one with a functioning app (minimalistic though it was), but apparently that was really beside the point.
Later I realized that my app, in truth, did not fall into machine learning category at all: experimentation with word2vec was by far not sufficient to qualify for this field. Realizing it made me laugh. If anything, the jury members were too generous to give it any points!
Thinking about it today for "no reason at all".
