VALUE PROPOSITION
- Increasing profitability and win ratio by generating analytically-supported insights
- Reducing participation in lengthy and costly lawsuits by improving case acceptance recommendations
- Decreasing labour-intensive tasks by providing a process-centric, AI-embedded application
Application of predictive modelling to class action lawsuits for a leading advisory firm specialising in litigation finance
BACKGROUND
The firm sought an analytically-based approach to class action lawsuit success and profitability forecasting
Legal experts struggled to identify and monitor all available information sources and incorporate all influential factors
Reliance on human judgement resulted in subjective decision-making and sub-optimal effort allocation
TARGETS
- Build a unified case knowledge base, seamlessly correlating information about the case, the defendants, and similar cases, across internal and external sources
- Convert masses of data into practical intelligence on the spot by connecting seemingly disparate pieces of data: parties, events, changes to financial reports, etc.
- Predict case results and financial outcomes based on all available data and past cases, making firm resources more impactful and profitable
ADDRESSING THE NEED
- Implemented an automatic questionnaire builder to compile all required information regarding a potential case, utilising Natural Language Processing and Entity Resolution (rationalising ambiguous data that correspond to real-world entities)
- Assembled a Knowledge Graph that captures a rich profile of class action lawsuits, automatically enhanced with information about the involved parties and events of interest from a variety of sources (e.g. market capitalisation of defendants, classes or types of securities involved, etc.)
- Assigned quantifiable scores and outcome estimates to potential cases using Pattern Matching, comparing Knowledge Graph attributes, relationships and effects to infer the most profitable cases