P2P

94 P E E R T O P E E R : I L T A ' S Q U A R T E R L Y M A G A Z I N E | W I N T E R 2 0 2 3 Data Assessment A common challenge that our clients in the legal field often encounter are the so-called "rolling uploads" – situations where processing work for a particular forensic collection is done on an on-going basis. Their primary concern is whether the machine learning algorithms will have to undergo multiple training rounds to recalibrate them for additional documents. Previously, the answer would depend on client's own estimates of how different the content or source of newly added data was compared to the existent sets. With a new feature that FRONTEO is introducing to KIBIT in 2023, the difference between data used to train AI and data that AI is being asked to evaluate can now be quantified numerically and KAM can produce an unequivocal answer as to whether additional sampling for training documents is required. There are other potential use cases for this technique – for example some clients have inquired about building a persistent data library of training sets for different types of litigation or investigation content. This functionality would enable a quick assessment of how effectively new data can be trained using a pre-existing training set from the library. Finally, FRONTEO is also working on utilizing this method for identifying the best or most optimal training set for a given document population, which would allow us to minimize the initial training process even further, driving further efficiency gains through the use of AI. Conclusion FRONTEO's flagship product – "KIBIT" AI engine – is a prime example of the company's commitment to accurate CJK data analysis. KIBIT relies on advanced NLP techniques to effectively identify and categorize words and phrases in multiple languages, including Chinese, Japanese, and Korean. The engine features several machine learning algorithms that receive continuous support and updates over time to ensure they are always able to deliver accurate results with minimum time investment. Moreover, FRONTEO's commitment to accurate AI analysis extends beyond its products and services. The company actively collaborates with academic institutions and industry organizations to advance research in AI and NLP technologies. This collaborative approach combined with FRONTEO's investment in advanced AI technologies and a robust team of expert developers and data scientists, all demonstrate the company's commitment to providing its customers with the most accurate and effective solutions to process, search and analyze CJK data. ILTA F E A T U R E S Vlad Lobatchev, an Engagement Manager and Director of Data Science and Strategy at FRONTEO, ensures that our clients always receive top-tier customer experience and best-in-breed technology to enhance their e-Discovery workf lows. He draws on his extensive professional experience and legal background to guide development and implementation of innovative technological solutions and workf lows on client engagements. Prior to joining FRONTEO, he was an attorney for two years specializing in general corporate and finance law – taxation, contracts, corporate formation, and administration.

• Cover