Products and Services
Overview
SENTIO's Continuous Active Learning method is analogous to a music streaming service "choosing" which songs a user will enjoy based on their previous song selections.
1
Users review and tag a small portion of the data set which is used to train SENTIO's prediction engine (a mathematical model) to order data from most to least (likely to be) relevant.
2
A model is built for each tag. With each new document tagged, the model becomes more accurate for each tag.
3
After a model reaches a certain quality measure, the model can be applied to the entire data set. SENTIO tracks the user tagging and determines when the training of the model is complete.
SENTIO sits between Processing and Review in the traditional eDiscovery workflow and relies on the text so it can handle all types of data, including email and office files. SENTIO offers real time Precision, Recall and F1 reporting. It shows the accuracy and completeness of the model being trained on-the-fly. Also, because the engine ranks the entire set rather than a limited set of randomly selected documents, SENTIO can integrate rolling document uploads into the process.
If later on new documents have to be added to the population after the training is finished, the previously trained model could possibly be applied to the new documents to generate results.
Review of Reviewers
Sentio Maestro can measure document review performance by utilizing Sentio's machine learning techniques to evaluate the accuracy and effectiveness of the reviewers at any given point in time. It can also identify responsive documents that may have been missed by individual reviewers with enhanced detection of false positives and false negatives.
