Our software engines leverage machine learning models and natural language processing (NLP) techniques assembled in proprietary algorithmic pipelines to mine insight from large, unconnected sources of text and structured data. Our models and NLP techniques are disruptive; they are built upon the leading-edge intellectual property of our co-founders and the scholarly work of UNC Charlotte’s Text Analytics Lab.
Our topic models use advanced neural embeddings to learn continuous concept vectors and then densify those vectors. These embeddings allows us to better capture local contextual information using less processing resources. We use sophisticated, domain tunable knowledge bases and improved text clustering engines that incorporate interactivity and multi-document summarization techniques to improve information retrieval accuracy and overall text clustering results. We are exploring the use of Mined Semantic Analysis as an improved tool for discovering latent concepts and enhancing the technology representations we mine from datasets.
Vector Analytics’ co-founders, Ms. Price and Dr. Zadrozny, have developed two patent-pending algorithmic pipelines that deliver predictive analytics for technology, product, and innovation benchmarking. These two pending patents utilize the technologies described above and are an early indication of the collaborative power of Vector Analytics’ co-founders.
Vector Analytics offers end users a view of what could happen tomorrow; i.e. we provide predictive analytics. Our competition only provides end users with data that describes what is happening today, i.e. they provide descriptive analytics.
Our software engines knit together elements of intelligence gathered from multiple, disconnected, unstructured, big datasets. Most of our competition offers analytics based on a dataset in singularly.
Vector Analytics is distinct from our competition because we do two things. We develop leading edge NLP tools AND we use those tools to provide end users with analytics. Most of our competition does one, or the other.