At Ozora Research, our goal is to build the world's most advanced tools for "cleantext" Natural Language Processing (NLP). Our work involves three steps:

  • Systematically developing sophisticated, knowledge-based technology for English text analysis
  • Sharing that knowledge with our clients through our APIs
  • Leveraging the technology to provide users with smart search capabilities

In contrast to most other NLP research, our strategy is to develop a deep linguistic understanding of English text, and then build that understanding into our software. Many people in the NLP world use machine learning algorithms to avoid having to learn about the structure of text - the algorithm learns for them. We use algorithms as tools to probe and investigate the structure of text, allowing us to increase our knowledge.