Inside the Human Language Project: A Local Data Collector's Journey
Interview with Kathleen Kownacki: Bridging Language Divides as a Local Data Collector.

In 2019, TAUS launched the Human Language Project, aiming to bridge the digital language gap for underserved communities. We work with contributors from around the world to generate, annotate, post-edit, evaluate text and speech data for various MT, ML and AI applications.

Today, we dive into the project's inner workings through an interview with a Local Data Collector for a Speech Data project, an initiative that harnesses human-powered data, focusing on speech data to create inclusive datasets for voice recognition and speech-based AI applications. Join us to explore the remarkable work of the Human Language Project and one of its dedicated Local Data Collectors, Kathleen Kownacki. Kathleen has several years’ experience in translation, along with a Bachelor of Arts in Linguistics and a passion for circus arts.

HLP: Was it the first time for you working as a Local Data Collector?

Kathleen: Yes, this was the first time I was in a role like this, but when I read through the description of the role and compared it to myself and my background, I thought I would be a good fit for the role.

HLP: How did you like the type of job?

K: I truly enjoyed it. The job not only ticked multiple boxes but also placed me in unfamiliar territory, which I could feel was helping me to grow. 

HLP: Was it a challenging role and why?

K: It definitely was, in very good ways. While it wasn't strictly a sales job, I often found myself reaching out to both acquaintances and strangers, trying to spark their interest. 

Figuring out how to market the project most effectively was a challenge though. So, building up contacts took a bit of work, particularly when I didn’t always get responses. I also had to sometimes navigate technical issues, like the occasional need to re-record speakers, which added to the challenge.

However, a unique challenge was assisting participants, especially those who might have dyslexia, through their recordings. I ended up employing strategies like following along with a finger or providing an audible preview, and I saw that those techniques tended to help them stay calm and accurate in recording, which was thrilling for me to see. It was very important to me that the process was accessible and an enjoyable and affirming experience for anyone who was participating.

HLP: How did you find the cooperation with the HLP team?

K: Throughout our cooperation, the HLP team was responsive and open to implement changes in their operations on the basis of my feedback. That helped tremendously in making future recordings easier, and I knew I was being listened to. 

HLP: What would your advice be for anyone interested in this type of role?

K: One bit of advice I'd have would be to spend the time it takes to make connections with key organizations. One breakthrough at one place can end up opening many doors. 

Somehow putting a personal touch on outreach can make a difference in my experience. For instance, I had sent an email and made a phone call or three to a wellness center director--all unanswered. But because I stopped by that wellness center in person again one afternoon, I ended up having a semi-chance meeting with that same director which turned things around once we'd had that personal contact face-to-face. 

Also, immerse yourself in what's going on where you're trying to recruit if you can. In my case, while I was waiting to talk to potential participants at a senior center, I participated in the senior fitness class that was going on, since I was allowed to have a table there to recruit during the class break. The energy in the class was awesome to me, and I saw that my willingness to participate with the seniors in the class helped to genuinely foster trust with some of them and brought more people to me and allowed them to be more open and willing to participate.

Lastly, don't underestimate promoting the project in more traditional ways, like with posters and flyers—they do work. It was from them that I got several willing participants who came to me


We are currently looking for a Local Data Collector in Australia. If you are interested, please check the job description on the HLP website.



Ariana is the Project Coordinator in the Human Language Project and Marketing departments at TAUS.

Related Articles
Participation in the language data projects offered by the Human Language Project leads to new professional opportunities.
Lexicala has emerged from the publishing world, as a provider of quality lexicographic content for leading dictionary publishers worldwide, and joined the TAUS Data Marketplace as a language data seller.
See how a medical doctor trying to break the linguistic monoculture in academia as a student ended up creating a dataset of medical glossaries and translation memories that brings about a 90% BLEU score improvement for the English and Arabic language pair.