Annotation done right.

Introducing Tagalog

High-quality labeled data is critical in artificial intelligence. Tagalog is our state-of-the-art solution for data management and labeling in Natural Language Processing.

Organize your datasets

NLP DATA MANAGEMENT BECOMES A BREEZE
Tagalog organizes all your NLP data, such as articles, emails, tweets and other document types in a list of corpora. You can prepare a corpus for labeling by adding annotation properties and labels, and reuse corpora for multiple applications.

Document classification, sequence labeling and information extraction

TAGALOG HANDLES ALL COMMON NLP TASKS
Tagalog’s labeling interface supports all common NLP tasks: single- and multi-label document classification, sequence labeling and information extraction. Shortcut keys help you choose between labels and jump to the next document quickly. Clear colors help you make fast decisions. Did you make an error? No problem: you can go back and update your annotation.

Single- and multi-label document classification

Sequence labeling

Raw texts, PDFs, scanned PDFs, JPGs and PNGs

TAGALOG PROCESSES A VARIETY OF DOCUMENT FORMATS WITHOUT PAINFUL CONVERSIONS
At NLP Town, we’re aware that texts come in many forms. In Tagalog, you can annotate simple text documents as well as multipage PDF documents (with their original layout), JPGs and PNGs. You don’t even need to split up your document collections: document types can be mixed within one corpus.

Seamless integration with popular OCR tools

IMAGES OR SCANS? NO PROBLEM!
Many people like to scan or take a picture of important documents, such as invoices or receipts. Even for these cases, Tagalog has you covered: it sends image documents to a high-quality OCR service like Google’s, and fetches the text for you to annotate. The flexible user interface lets your labelers work efficiently by selecting words or entire blocks of text in the original image.

Sequence labeling on images with OCR

Multi-user environment

LABELING IS A TEAM EFFORT
High-quality, large-scale data annotation requires collaboration between a team of annotators. As a project manager, you can have multiple annotators work on the same corpus simultaneously. Optimize for speed by splitting up the corpus between annotators, or for quality by requiring all annotators to label all documents. Tagalog's management panel keeps you up-to-date with the progress of your annotators and helps you monitor annotation quality with the inter-annotator agreement matrix.

Labeling campaigns

ORGANIZE LABELING PROJECTS IN CAMPAIGNS AND TASKS
Tagalog lets team members work together in labeling campaigns. As a project manager, you set up your campaigns, assign annotators and reviewers to the campaign tasks, and follow their progress in real-time. Even during the campaign, you can relocate annotators or reviewers to keep your project on track.

Side-by-side comparison reviewing

Quality assurance

REVIEW ANNOTATIONS SIDE-BY-SIDE
As a reviewer, you assure annotation quality by comparing the inputs from multiple annotators side-by-side. You can accept correct annotations or make corrections where needed. Tagalog helps you work effectively by highlighting non-matching labels in texts, images, (multi-page) pdfs, etc.

Want to know more?

Would you like to find out more about Tagalog? Fill in the form below and we’ll get in touch.