How to make data labeling for free?

Who?

ML in PL is one of the most recognizable associations in Poland working to promote machine learning and methods of understanding it. Them organized the most recognizable conference for students in the field of machine learning. Thanks to ML in PL that the DataLabeling.EU team had the opportunity to meet and carry out the project with Alphamoon.

It all started with a competition organized by ML in PL in 2019. The Alphamoon team won 70 hours of free text data labeling where WEimpact.Ai and the DataLabeling.EU brand were sponsors.

ML in PL Conference - wpis gratulacyjny dla zwycięzów.webp

Zespół Alphamoon na konferencji ML in PL.webp

Alphamoon is a Polish company dealing with AI solutions for business. They specializes in the areas related to Natural Language Processing and Computer Vision, offering their solutions on the international market.

What have we done?

The first step covered an overview of the scope of tasks, amount of data and lead time as well as detailed labeling instructions. Then our coordinators performed tests to verify the time-consuming nature of the entrusted tasks and the initial quality of their performance. Four annotators were signify about 1000 pages of text in a few days, accurately labeling individual elements in subsequent stages (image with markings below).

Obraz z oznaczeniami etykietowanych danych.webp

Effects:

As a result the Alphamoon team gained tagged data in less time for their model whose task is to recognize visually distinctive elements on the page.

Opinion on cooperation:

We evaluate the cooperation with DataLabeling.EU as very good. The company is characterized by a professional approach in terms of contact with the client as well as efficient and timely performance of tasks. Any comments we had were immediately taken into account. We especially value their professionalism and commitment to the commissioned project.

Karolina Cwojdzińska Junior Machine Learning Engineer @ Alphamoon.webp

Karolina Cwojdzińska — Junior Machine Learning Engineer at Alphamoon

Aneta WróbelCOO w WEimpact.Ai | Koordynator projektów etykietowania danych

Posiada bogate doświadczenie w zarządzaniu projektami etykietowania danych i koordynacji zespołów. Specjalizuje się w nadzorze nad projektami anotacji danych głosowych, językowych i obrazowych, co jest kluczowe dla rozwoju technologii AI. Jej ekspertyza obejmuje optymalizację procesów, zarządzanie zasobami oraz zapewnienie wysokiej jakości danych treningowych dla modeli uczenia maszynowego.