About Me

My name is Ricardo Lezama, I am a Chicano linguist based out of California and the tri-state area (New York, New Jersey and Philadelphia). At any given moment, I can be found in any of those states, but, in my heart, I am a Californian. And, a linguist. I also love to visit as many cities as possible to open myself up to new environments and their ideas. Corny, yes, but I have a right to be since I worked extremely hard to get what I have.

Many of the resources here are intended for humanities majors looking to grow in computational expertise. My advise is for you to not try to become computer scientists. Consult with them, but do not compete. Instead, become a computational version of your discipline because that is both infinitely more interesting – given your vantage point – and viable. Example: “Computational <your_discipline>”. I am certain that if you run a search for these two terms, then you will be led to accessible materials for you to sharpen up relevant computing skills. Anyone advising you to fight defensively (go back and do XYZ course) is advising you to play catch-up (I have some thoughts on this point) in a race you likely already lost. Instead, play to your strengths.

Me in Austin, Tx on 5/16/2021

Like for most Linguists, business decisions always involve a comprimise between what you as an individual were trained to do and what your environment lays before you. Thus, I have evolved into a ‘Computational Linguist’ professionally since my first gig at Intel in 2013. On a broad level, my work intersects fine grained language generalizations and technology. The latter for me means Python, open source libraries for NLP and a lot of Keras, lately. I recommend you try to leverage Spacy whenever you have an interesting project.

I currently work in Bloomberg LP as part of the News Indexing team. That News Indexing product is likely the best news classification engine in the world, with Reuters likely being a second or alternate candidate for that 1 spot. I don’t have the information to evaluate that they are, though, so, I stick to Bloomberg’s News Classification product being the top dog.

On a daily basis, I may analyze large amounts of text to understand what search terms may best reference them. I also explain my findings to people whose day to day is very different from mine. In other words, some of your work can have a scripting component, but also an explanatory one. Do not neglect scientific (but very abbreviated) style explanations. They are majorly useful in a place where time is literally quantified as money.