High-Precision Biomedical Relation Extraction for Reducing Human Curation Efforts in Industrial Applications

Exploiting graph kernels for high performance biomedical relation extraction

BACKGROUND Relation extraction from biomedical publications is an important task in the area of semantic mining of text. Kernel methods for supervised relation extraction are often preferred over manual feature engineering methods, when classifying highly ordered structures such as trees and graphs obtained from syntactic parsing of a sentence. Tree kernels such as the Subset Tree Kernel and Pa...

Kernel-based learning for biomedical relation extraction

Synthesis of Zataria Multiflora extraction loaded chitosan/PEO nanofibers via electrospinning method for potential biomedical applications

Background & Objective: The nanofibers are referred to as fibers with a diameter of less than 1000 nm. There are various ways to make nanofibers, one of which is electrospinning. In this method, nanofibers are formed on aluminum sheets. The conversion of chitosan to derivatives with greater solubility, its mixing with other polymers and the use of special solvents are methods for its ease of us...

Dynamics analysis of microparticles in inertial microfluidics for biomedical applications

Inertial microfluidics-based devices have recently attracted much interest and attention due to their simple structure, high throughput, fast processing and low cost. They have been utilised in a wide range of applications in microtechnology, especially for sorting and separating microparticles. This novel class of microfluidics-based devices works based on intrinsic forces, which cause micropa...

High Precision Information Extraction

ABSTRACT Most fully automatic information extraction systems achieve less than 100% extraction precision and recall. On real applications these parameters typically vary between 50% to 95%, depending on the extraction method and source data. We present an information extraction system designed for applications where little or no error can be tolerated. The system is not fully automatic. Instead...