Machine Learning for Biomedical Literature Triage

Databases allows storing data in a consistent way, facilitating easy retrieval and enabling both complex searches and computation on data. In the biomedical field, databases are also used as vital resources for scientists searching literature. Over the past few years, researchers and users have noted a significant expansion of such literature databases [1]. For example, the free on-line database PubMed [2] currently holds over 22 million documents, and a simple keyword search can retrieve more than hundreds of thousands of documents. The analysis of the vast biomedical data currently available is a challenge addressed by studies such as [5] [6], as well as the use of this data to identify relevant information for biomedical research [7] [8].