Predictive Filtering for Electronic Discovery Document Review
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In electronic discovery systems, the sheer volume of documents makes it impractical for reviewers to manually analyze each document to identify relevant or responsive documents, and existing coding methods lack context to determine the significance of tagged documents.
Innovation Solution
The system employs predictive filtering by generating automated filter values based on predictive modeling, allowing for the identification of responsive phrases, concepts, and meta-data, which are then used to suggest filter criteria and refine document searches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual document analysis is performed by reviewers, then context and significance of tagged documents can be determined, but the sheer volume of documents makes it impractical to analyze each document
Solution Approach 1:
The system performs preliminary automated coding of documents with predictive values before human review. Reviewers receive pre-coded documents with predicted responsiveness and key phrases already identified, allowing them to focus on reviewing the significance of pre-identified responsive documents rather than analyzing every document from scratch.
Solution Approach 2:
An automated predictive coding system acts as an intermediary between the large volume of documents and human reviewers. The system generates predictive values, key phrases, and responsiveness predictions that serve as intermediate artifacts, enabling reviewers to efficiently identify significant documents without manually analyzing each one.
2Adaptability or versatility
If existing coding methods are used to tag documents, then documents can be organized and searched, but the methods lack context to determine the significance of tagged documents
Solution Approach 1:
The system applies different levels of coding detail to different documents based on their characteristics. Rather than uniform coding, reviewers can drill down from high-level responsiveness predictions to specific key phrases and concepts for individual documents, providing localized contextual information where needed.
Solution Approach 2:
The system adds a new dimension of predictive values beyond traditional binary responsive/non-responsive coding. Multiple predictive values including key phrases, concepts, and confidence scores create a multi-dimensional coding framework that preserves contextual significance while enabling versatile filtering and searching.
3Productivity
If predictive filtering is implemented to reduce document analysis, then review efficiency improves, but the system complexity increases
Solution Approach 1:
The predictive coding system is divided into separate functional modules: document analysis components that generate predictive values, key phrase extraction components, confidence scoring components, and reviewer interface components. This segmentation allows each module to be independently optimized and maintained, reducing overall system complexity while maintaining high productivity.
Data Source
AI summary
Electronic discovery using predictive filtering is disclosed herein. An example method includes receiving a plurality of documents. A selection of a filter value is received, where the filter value comprises a field value or a set of field values for the plurality of documents. A prediction of responsive phrases, responsive concepts, or other meta-data is generated. The plurality of documents is evaluated based on a new filter value. Then, a prediction of other responsive phrases or other responsive concepts is generated. A predictive value is generated. Filter criteria is generated based on the predictive value for each of the other responsive phrases, the other predicted responsive concepts, or other predictive meta-data. A selection is then received of the filter criteria. A filter is built and applied based on the selection, and documents are generated from at least a sub-portion of the received plurality of documents.