Semantic similarity word game powered by Rosette
See how one Hackathon team used Rosette to create a multilingual word game based on semantic similarity.
Learn moreReleases
NEW Release 1.23 (September 2022)
NEW Release 1.22
NEW Release 1.20
NEW Release 1.19.3
Improved visual annotation data model makes it easier to compare entity extractions
See how one Hackathon team created an annotation data model integrated with Rosette for easier comparison of entities.
Learn moreTackling the entity disambiguation challenge: a new approach shows promise
New research from BasisTech may improve entity disambiguation. It discusses ways for software to more easily recognize and disambiguate entities appearing in disparate data streams.
Learn moreWhat is lemmatization? Learn why this process is vital to language processing
Lemmatization — or the process of eliminating prefixes, suffixes and compounds from words to find their most essential forms — enables comprehensive searches and automatic extraction of themes from text.
Learn moreCelebrating National Crossword Puzzle Day
BasisTech celebrates National Crossword Puzzle Day on Dec 21 with our own AI and linguistics puzzle. Learn more about using AI to solve puzzles.
Learn moreHoney, everything’s coming up Rosette!
Rosette shows up on vacation, in ways both precise and fuzzy ways, just like our text analytics do!
Learn moreWhat is fuzzy matching? Explore how fuzzy logic boosts name-matching accuracy
By taking into account real-world issues such as typos, misspellings, and disordered data, fuzzy name matching improves on exact matching processes.
Learn more
Koat.ai
Online conversations are muddied by noise from narrative manipulators, including marketing campaigns and traditional media. koat (koat.ai) has found a way to count the voices of real people and online ...
Learn moreName-matching fails: The lighter side
From the toddler “terrorist” to an official named Mícheál not “Lord Mayor,” we examine some humorous name-matching fails.
Learn moreName matching fails: the real-world consequences
Imprecise name matching processes wreak havoc in finance, law enforcement, healthcare, and other sectors.
Learn morePattern-based few-shot entity disambiguation reduces need for large data sets
Learn about exciting new approaches for few-shot learning that can make powerful entity disambiguation models with a fraction of the training.
Learn moreWhat is transliteration? Why does it matter? Take an up-close look.
What is transliteration? The process of systematically converting text from one script to another in predictable ways is vitally important to national security, finance, and other industries.
Learn more
DOKS
In Europe, businesses from the largest financial institutions to individual bookkeepers, attorneys, and real estate brokers must identify their customers to comply with European anti-money laundering (AML) screening regulations. These ...
Learn moreAI in finance helps institutions combat threats, improve compliance
In a global climate marked by political unrest and evolving technology, AI in finance helps institutions reduce crime and improve compliance.
Learn more