We are the Text Analysis and Knowledge Engineering Lab, a research group at the University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia. We specialize in natural language processing, machine learning, and text analysis.
We apply state-of-the-art machine learning and natural language processing techniques to help people interpret and make sense of vast quantities of text data.
No, it is not a dog in our lab 🐶.
TakeLab Retriever is a platform that scans articles and their metadata from Croatian news outlets and does text mining in real-time (Yes! Articles are being processed and semantically analyzed as you are reading this!). We have already analyzed over eight millions of articles, and analyze 15000 new articles every day.
To obtain access, please fill out this form at https://forms.gle/UC8jk2u2FaL4g3JP8. We will review your request as soon as possible.
Retriever allows you to analyze the articles using advanced text-searching mechanisms powered by machine learning and natural language processing. You can sift through Croatian news outlets using phrases, organizations, people, locations, and much more. Reveal trends, patterns, and correlations in the outlets with just a few clicks and keystrokes.
Retriever will analyze the articles from the outlets you’ve chosen and show you the result of your query. These articles will be presented graphically and in a table. The graph view is designed to reveal trends and correlations in articles over time. The table view is a reflection of the graph view. The titles of articles that make the graph are shown alongside their metadata in the table. You can also download the table snapshot at any time if you want to get the articles' list and their metadata (URL, title, date of publication, etc.). To obtain the original text, visit the official news outlet website.
While Retriever is primarily intended for academics (such as political scientists, media analysts, psychologists, and sociologists) who wish to analyze media thoroughly and painlessly in just a few clicks, it can be used by anyone with non-commercial interests. TakeLab Retriever is a window into the present and history of online Croatian media. As such, it is a must-use for researchers who want to find trends, patterns, and correlations, which are otherwise inscrutable when searching with other, general-purpose search engines, such as Google, Bing, or Yahoo.
So off you go, stop reading this about page, and use our explorer to find out how your favorite Croatian prime minister coincides with buying warplanes or how the media views the famous singer when he or she shares fake news. Or maybe you are interested in how a specific phrase or a person appears in left or right-wing outlets. That is up to you. Your queries are never saved, and we never track you. The possibilities are endless and are bound only by your wit and imagination 🪄.
Good luck, and if you have any questions or suggestions on how to improve Retriever, drop us an e-mail at retriever.takelab@fer.hr ✉️.
MLA Dukić, David, et al. "TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets." arXiv preprint arXiv:2411.19718 (2024).
APA Dukić, D., Petričević, M., Ćurković, S., & Šnajder, J. (2024). TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets. arXiv preprint arXiv:2411.19718.
CHICAGO Dukić, David, Marin Petričević, Sven Ćurković, and Jan Šnajder. "TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets." arXiv preprint arXiv:2411.19718 (2024).
HARVARD Dukić, D., Petričević, M., Ćurković, S. and Šnajder, J., 2024. TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets. arXiv preprint arXiv:2411.19718.
VANCOUVER Dukić D, Petričević M, Ćurković S, Šnajder J. TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets. arXiv preprint arXiv:2411.19718. 2024 Nov 29.
BIBTEX
@article{dukic2024takelab, title={TakeLab Retriever: AI-Driven Search Engine for Articles from Croatian News Outlets}, author={Duki{\'c}, David and Petri{\v{c}}evi{\'c}, Marin and {\'C}urkovi{\'c}, Sven and {\v{S}}najder, Jan}, journal={arXiv preprint arXiv:2411.19718}, year={2024} }