Looking for something specific beyond the first page? How about something relatively unknown or not searched by many? Or, what if you lack trained data, models or resources for AI, and Machine Learning based Search Engines?
Customers such as biologists, reagents suppliers constantly struggle with limitations of general purpose search engines or do not have resources to build one for their specific needs.
A brand new experience - Color coded graphs, filters for topics, individual concepts, how they relate to one-another, and their references.
Self Learning - A system that continually learns from search context, sentence structure, dictionaries, catalogs and explicit user feedback on search results.
And Highly Customizable - build new UI, add new data, apply new dictionaries, tweak engine parameters, or try out new algorithms.
We present results as a collection of interrelated concepts that are grouped together in various topics - such as products, places, events, people and others that are dynamically generated based on context of search terms.
These concepts are related to one another by their common references in source documents. More frequently they appear together, stronger their relationship is. These relationships can also be typed such as pathways, transformations or signaling based on how concepts are described together in a sentence.
NLPCORE collects statistics such as word frequencies, colocation frequencies (within a single document and across the corpus) along with part of speech tags (noun, pronoun, or verb) in its index. Words that appear most frequently and closest to the search term(s) provide the seed data set for neural network algorithms that also factor in any available heuristics, dictionaries or past user feedback to present the most appropriate concepts and their relationships.
Our entire search platform from crawler, parser, indexer, data store, dictionary lookup to neural network search, is built upon discrete web scale components that are independently deployed, accessed through well defined APIs and run at scale for optimum throughput.
We help researchers to:
* Identify candidate bio-entities, reagents, and their specific interactions for their experiments
* Validate references against biological dictionaries and published research
* Provide inline feedback to improve results accuracy
We help manufacturers and suppliers of life sciences materials to:
* Track usage of their products in research together with other related products
* Identify new products created or discovered in research
* Identify competitive products in use
NLPCORE is architected across independently scalable components with extensibility at its core. Our secure APIs help:
* Solution Developers build new vertical solutions - Legal, News, Product Support, Resumes/Job Search
* Data Scientists aggregate results as time series, tag cloud, or geo map, and
*Algorithm developers apply new algorithms on raw results, index data with minimal development effort.
Visit our APIs page to learn more.
Center for Infectious Disease Research (CIDR)
Center for Infectious Disease Research is the largest independent, non-profit organization in the U.S. focused solely on infectious disease research with the mission to make transformative scientific advancements that lead to the prevention and treatment of infectious diseases.
University of Washington (UW)
University of Washington is a leading university in the world known for its research in medicine, science, as well as its highly-competitive computer science and engineering programs. We are working together to provide scientists easy to use/reuse and easy to deploy tools for their bioinformatics workflows.