tar | INTELLECTUAL DATA
Data Analytics
Technology Supporting AI-driven
Document Review to Save Time and Cost
Data Analytics TAR (Technology Assisted Review)

This is a technology used for document review leveraging AI.
Based on a Human Reviewer’s document review results, AI learns
to determine relevance Yes/No, and organizes the documents
determined to have higher relevance to be prioritized for review.

  • Real-time AI Analytics

  • Accessible to Various Project

  • 01
    Categorization and Review
    of Human Reviewer’s Documents

    Out of the immense number of documents,
    reviews documents determined to have higher
    relevance and categorizes the other
    documents to narrow the search.

  • 02
    Real-time Learning by AI of
    Human Reviewer’s Review Results

    Based on the insight of experts
    received as input, AI conducts analysis
    and separation of the remaining data.

  • 03
    Fast and Accurate
    Access

    Through real-time learning, AI analyzes
    which materials are needed by the Human
    Reviewers and prioritizes them according
    to relevance.

What is Active Learning
Implementing TAR?
Machine Learning Technology Capable of Active Learning

Active Learning is a unique part of machine learning. It communicates with the Human Reviewer who provides input for the AI to learn, and proactively drives towards the point of decision-making. It mitigates uncertainties of the outcome and increases the accuracy of matters that are relevant. As a result, the outcome is delivered with speed and accuracy.

How to apply Active Learning (AL)

By leveraging Active Learning, documents can be analyzed in real-time,
and Human Reviewers can receive support on prioritizing documents of higher relevance for review.

AL = SVM Learning Support Vector Machine

This is a type of machine learning technique using Binary Classification to determine the range of a specific data set.
After learning the patterns between the data sets during training, decision-making is executed based on the support vectors of the two groups.
When there are ambiguous boundaries, the system takes a proactive approach to make a determination.

  • Machine Learning
  • Pattern Recognition
  • Determination of Ambiguos Boundaries
  • Effective Learning Capability