How can DoCTER help you?
DoCTER, ICF's Document Classification and Topic Extraction Resource, is a free web-based software application that helps explore and prioritize documents — such as abstracts from the scientific literature or social media messages — for expert review. Using algorithms from the domains of natural language processing and machine learning, DoCTER improves efficiency in tasks that involve large volumes of text.
What does DoCTER do?
DoCTER uses two alternative approaches to automated document classification and prioritization: machine learning and supervised clustering. DoCTER also has a topic extraction or clustering function that helps identify the principal themes in a body of literature. More detail about when and how these functions are best deployed is available in the Tutorials section.
How do users interact with DoCTER?
Users first choose the DoCTER function they wish to deploy. Text to be analyzed must be provided to DoCTER as a csv file in a prescribed format that will be uploaded from the user's computer. Parameters for the analysis must be specified and submitted in browser forms as prompted. Following cloud-based analysis, output will be available for download to the user's computer. Tutorials and sample input files are provided to illustrate the application's usage.
DoCTER in Publications
Supervised clustering for automated document classification and prioritization: a case study.
View details »
Conference Presentations
Simulating time saved for manual screening of literature for IRIS assessments using text analytics.
View details »
Blogs and White Papers
Machine learning could make document classification easier and more accurate — if we deploy it correctly.
View details »