Crowbird Spotlight: Adriana Picoral

Crowbird Adriana Picoral is a prime example of taking an interdisciplinary approach to academic research. Passionate about computer coding since the age of nine, Adriana always knew she wanted to be a computer scientist. Unfortunately, with female Computer Science students outnumbered by a ratio of 1 to 15 at her university (Federal University of Rio Grande do Sul, in Brazil), Adriana’s presence in a STEM-focused major was constantly called into question. Jokingly, she credits her eventual interest in linguistics research to “running away from computer science because they were mean.” In reality, Adriana’s undergraduate thesis on developing a computer game to teach Portuguese to non-native adults is what sparked her interest in language learning.

Adriana’s research process has come a long way since her undergraduate thesis, but one key element has remained the same: a focus on interdisciplinary methods and tools to understand language acquisition. Her research analyzes the intersection of corpus linguistics, computational linguistics, and foreign language acquisition. For her dissertation, Adriana is researching how different factors affect third-language acquisition in adult learners. Specifically, she is looking at Spanish-English bilingual adults, and investigating how their native language affects their ability to learn Portuguese. She uses mixed methods by creating a corpora of Portuguese, English, and Spanish texts and then applying computational linguistics methods to analyze the language behavior.

Graphic used in Adriana's dissertation on copula verbs to adverbs
Preference of ESTAR copula use with intensifiers across different corpora for Adriana’s dissertation

But as much as she enjoys research, Adriana isn’t ruling out the possibility of working in industry instead of academia. In her internship with the Educational Testing Services (ETS), Adriana discovered how valuable an interdisciplinary researcher is in an industry already saturated with specialized employees. This became further evident in her 2018 internship with Google, where there was an abundance of linguists and software engineers, but not many employees who could do both, like Adriana.

After taking a corpus-linguistics class taught by Crow co-founder Shelley Staples, Adriana became a Crowbird in the fall of 2016. Since then, she has put her computer skills to work by standardizing the text format of Crow’s collected materials. Using her experience in coding, she worked with Shelley to create a system that converts all documents to normalized UTF-8 text files. This also labels the words in the texts for speech tags, such as verbs or nouns, for future language analyses. Adriana also created Python scripts to ensure repository materials are tagged and encoded accurately, and JavaScript web-interfaces to assist in manually coding students’ texts for a number of things, such as citation practices.

Aside from her work doing text nominalization, Adriana has also participated in multiple Crow workshops. In July 2018, she helped lead the debut of the Crow web interface in a 3-hour workshop at the Teaching and Language Corpora (TaLC) conference in Cambridge, England. That same year, she presented a comparative analysis of various linguistic tagging tools at the 14th American Association of Corpus Linguistics conference and a workshop on the citation practices of L2 writers at the American Association for Applied Linguistics (AAAL) conference.

Adriana with colleagues at AAAL 2019
Adriana (back, second from right) with colleagues at AAAL 2019

Moving forward, Adriana is interested in taking Crow’s research on citation a step farther by incorporating the computational methods she used in her dissertation into Crow. She intends to create machine learning models to classify new data. She is excited to work on a project that unites Crow work with her dissertation research. The ability to incorporate different interdisciplinary approaches into her work is Adriana’s favorite part about Crow.

We look forward to seeing how Adriana will continue to improve our interface and promote interdisciplinary research methods.