Literature Mining of Antibiotic Resistance


Antibiotic resistance is a very important part of human health that needs to be constantly kept track of in order to give proper medication to patients. However, the large number of papers that come out makes it very difficult for a human to keep up. Thus, this project aims to create an automated way to extract meaningful information out of antibiotic resistance papers. It aims to create an automated way to collect and parse the papers. Then, it tries to analyze the papers and extract meaningful information out of it. In order to do so the project was divided into 4 main parts. In the first part antibiotic resistance genes were gathered to be used as search queries. In the second part PubMed papers were gathered using the genes. In the third part DeepEventMine was used to extract information from the papers. In the last part the extracted information was analyzed. The results from the data gathering process were good and lots of relevant papers were gathered. DeepEventMine was run successfully but most of the output was not very useful. The analysis part was mostly focused on statistical analysis and gave some useful results. As for the final deliverable the code used for data collection works well and can be used on other projects that work with medical papers. Detailed information on how to work with DeepEventMine can be found in this report. Finally, more work can be done in the analysis section to produce more information.

antibiotic resistance, data mining, event extraction