Santa and Sentiment Analysis on Christmas

  2 m, 11 s

Every Christmas Eve, I visit 700,000,000 children around the world to deliver their presents. But the complexity of each visit varies greatly, depending on the behavior-to-date of the child in question. I generalize this behavior into two buckets: Naughty and Nice.

Now, Naughty and Nice unfold into their own taxonomies, which help me gain a more granular view of each child’s behavior in context. These taxonomies are, of course, pro-ho-ho-ho-prietary. But I’m not all secrets! I base my behavior analysis on language data, which I find to be the most straight-forward and exhaustive source. Mind you, this doesn’t only account for the child’s language data, but also the language data of their parents and guardians! When Natasha is scolded by her aunt, that intensifies the weight of naughtiness in her overall behavior score.

Now, parents or guardians aside, consider for a moment that (by my calculation) the average child speaks 3,500,000 words a year. Multiply by 700 million, and that leaves us with a lot of natural language data to work with! When I’m making my List, and and while my team of Data Scientist Elves are checking it twice, we must make sure to emphasize both precision and recall.

For years, my elves had to take time off from making toys to build and maintain an open-source solution for List checking. And when parameters went out of whack, it took even more attention away from toy-making. A great example of where this got out of hand was in 1912, during The Great North Pole Rocking Horse Crisis. A whole cohort of elves needed to tune our configuration, which failed to adjust to the changing vernacular of the day. This led to rocking horse production falling by 11%!

In the North Pole, optimization is everything. That’s why I decided to partner with Lexalytics. Their AI-powered natural language processing solutions make it a breeze to check the List twice. Their language packs mean I never need to deal with sloppy translations. And the support they offer is second to none.  What’s more, their trial period was as enjoyable as roasting chestnuts on an open fire.

Using Lexalytics for sentiment analysis allows me to make informed predictions for the sleigh plan, and knowing with certainty who will get coal and who will get presents lets me optimize my route. The best part of having it all planned in advance: I get to spend more time eating cookies! And on that note, I must to see to the buffing of the sleigh runners. But, I’ll be seeing you all soon.

Happy Christmas to all, and to all a good night!


Categories: Sentiment Analysis, Special Interest