“Success is no accident. It is hard work, perseverance, learning, studying, sacrifice and most of all, love of what you are doing or learning to do.” ― Pele, 1940-Present
Matthew Marting, working in collaboration with Apertium, won one of the coveted Grand Prize Winner slots in the 2016 Google Code-in. The winners are flown to Google campus for four days in the summer to meet with Google Engineers and hang-out in Silicon Valley. Matthew completed over 20 tasks with Apertium in the years he worked on their open source project using multiple programming languages and learning the fundamentals of language, syntax, and translation along the way.
To compete in the Google Code-in competitions, teens pair up with one or more of the 17 open source organizations. In cases like Apertium, these tasks can be non-trivial and require professional level skills to accomplish. This is where the learning and learning curve come in – in order to accomplish the tasks, young coders like Matthew have to learn about the existing code base, the objectives of the open source organization, and the coding language and methodology to even start making a meaningful contribution. On average, the Grand Prize Winners completed 25 tasks that contributed to code in the open source community – some of which is at the heart of things that we use and take for granted everyday. Matthew’s work is an example of such compelling work and contribution.
Matthew graduated from St. David’s School (the Warriors) in Raleigh and is now an incoming freshmen at Georgia Tech in Atlanta. He has also been an intern at Device Solutions in the summers of 2016 and 2017. As it turns out, the work that Matthew did to win the Google Code-in and the need to learn new coding language(s) for work at Device Solutions are closely coupled. Matthew started his Code-in journey back in the 9th grade and has participated every year – a senior at St. David’s School was the first person to mention the competition to him. Matthew selected some of Apertium’s machine language translation platform tasks to cut his teeth. The first year that Matthew worked with his assigned mentor at Apertium he was included as a co-author on a published paper based on tasks where he had contributed and advanced their code base (errors in monolingual dictionaries, bilingual dictionaries, and basic lemmas) – talk about street cred!
When you dig in deep with Matthew about what he actually worked on, you swim in words like “parser”, “transfer rules”, “translation pairs”, “lemmas”, “algorithm”, “finite state transducers”, “bigram tagger”, “linearization”, “memory leaks”, “Debian”, “Ubuntu”, and many many more. Whether it is coding with C++, Python, XSLT, or Java, Matthew has a reasonably good handle on them all – all a result from his determinism to understand the existing code, its limitations, and where he needed to concentrate his time to make improvements. In summer before and during 11th grade, it was determined that in order to clean some things up he was going to have to completely rewrite certain modules along the chain – most specifically the tagger and parser on the front end. To get all this work done, he spent many long days during the year and on school breaks – serious dedication from a high school student.
The experience with Apertium and Google Code-in has led Matthew to greater competence and confidence in coding with both local and virtual teams, not to mention it has provided great content for college essays with Apertium serving as a great reference. When you consider that all of the work Matthew did for Apertium was volunteered time it makes the Grand Prize even more impressive. When asked why he did it, Matthew responds very nonchalantly: “It seemed like a cool thing to do.”
For more on Google Code-in: https://opensource.googleblog.com/2017/01/announcing-google-code-in-2016-winners.html