Gather all the information from every single document.

TagWorks was designed by a social scientist to analyze large sets of documents using complex conceptual schemas. This web-based system can help you finish your giant data labeling project up to ten times faster.

ui graphic-site.png

What can TagWorks do?


Complexity at scale

With TagWorks you can tackle complex, large-scale projects with ease. Efficiently annotate, tag and classify tens of thousands or millions of documents with hundreds of categories of information.


Up to 10x faster

Complete what would normally be a decade long project in a year. Tagworks removes the need to train wave-after-wave  of research assistants, saving you time and money.


The power of the crowd

TagWorks allows you to easily enlist thousands of crowd workers to extract the information you need. In most cases, anyone who can pass a reading comprehension test can contribute.


Eliminates task management

TagWorks automates worker, document, and task management, allowing Principle Investigators to principally investigate instead of spending so much time training team members and directing traffic.


Validated results

TagWorks is open and transparent, making it simple for others to review and validate your results. You can easily see the full provenance of your data and share it with your peers.


Web based

TagWorks is completely web based, so no software installation or management is required. This also allows crowd workers and other collaborators to have easy access to your project.


How does TagWorks work?

  1. Gather your corpus of documents and specify what you want to find into a conceptual scheme.

  2. The TagWorks team works with you to convert your conceptual scheme into a branching hierarchy. Next we create an assembly line of tasks guiding the annotation efforts of online crowd workers.

  3. Upload your documents into TagWorks.

  4. Expert members of your team perform TagWorks-generated tasks for your documents. This work will establish a “gold standard” set of high-quality tags that TagWorks will use to to test and qualify crowd workers.

  5. Open the starting gates and introduce your thousands of documents to thousands of crowd workers or volunteer citizen-scientists.

  6. TagWorks automatically compares each crowd worker’s tags to at least three others’, formalizing and exploiting intersubjective epistemology to produce validated results.


Who uses TagWorks?


Columbia University’s History Lab is using TagWorks to annotate an archive of 1,000,000 diplomatic reports.


The University of Texas School of Information is using TagWorks to categorize mentions of open source software in thousands of articles.


Public Editor is using TagWorks to transparently assess the credibility of news articles and news organizations with the help of thousands of volunteers.


Where did TagWorks come from?


Sociologist Nick Adams wanted to better understand the complex social interactions during the Occupy movement. The research scrutinized thousands of events across hundreds of locations. He began looking for labels of sociological concepts, then organizing into time series and multi-level models. In the end his team annotated over 8,000 articles with about 300 variables. The result was a rich dataset that gave deep insight into police and protester interactions.


The process was lengthy and involved many research assistants. Adams worried that conducting similar research would be at a high cost and require extensive supervision.
There was no software at the time that could handle the nuance of these sociological concepts.

Many ambitious research attempts have faced similar challenges. Those projects had either simplified their approach or reduced their document set.


Adams reached a breakthrough when he began to combine hand-coded analysis with automation. He broke down the process into simple tasks that require little training. This new assembly line approach led to the creation of TagWorks.

Adams teamed up with veteran software engineer Norman Gilmore. Together they work to bring this solution to other projects seeking to answer big questions.


 When Is TagWorks The Right Solution?


When it is

If you have a large (or even gigantic) corpus and/or you have an intricate conceptual scheme with dozens or even hundreds of classifications, TagWorks may be the solution you are looking for.

When it isn’t

If you have less than 500 documents to analyze and/or you have a simple conceptual scheme with a handful of categories, there are a number of other tools that can meet your needs.  


If you’d like to learn more about the suitability of TagWorks for your project, get in touch and we’ll set up a free consultation.


Get started

Interested in using TagWorks? Complete the form below and we’ll chat soon to schedule a free consultation.

Name *
Join our mailing list