We're excited to announce the release of our new data programming feature. It's no longer restricted to row-based projects - you can now use it for all token-based projects too! This means you candeploy weakly supervised learning on your NER or POS workflows, for example.
In traditional labeling, humans have to manually label the data, which is a time-consuming and expensive process. Data programming uses algorithms to automatically and accurately label data. Datasaur leverages weak supervision, which has made the process even more efficient and up to 9.6x more effective than manual labeling.
Weak supervision is a technique used to automatically label data using multiple, often noisy, sources of supervision. The goal of weak supervision is to reduce the time and cost required for data labeling while still maintaining a high level of accuracy. This technique is particularly useful in cases where there is a large amount of unstructured data that needs to be labeled.
Here is how you can use data programming for your Row-based projects. As a reminder, row-based projects are textual classifications workflows (e.g. -sentiment analysis). Additionally here is how you can use data programming for Token-based projects.
For more information, please refer to our documentation here or contact support (firstname.lastname@example.org)
We're currently working on a no-code version of our data-programming feature! This will allow even more users to take advantage of the power and efficiency of data programming without the need for any coding knowledge. Stay tuned for updates on our progress and release date.
To get started with our data programming feature, please contact our support team (email@example.com) to request access. They will be happy to assist you with any questions you may have and help you get up and running with this powerful tool.