COVID-RAPID-MINER; Information summarizer
Rapid rise and spread of the severe acute respiratory syndrome coronavirus (SARS-CoV-2) across the globe put crushing demands on health-care workers and hospital resources. Not sufficiently prepared for a pandemic as large as COVID-19, drugs, vaccines and other therapies are not yet readily available. Nevertheless, in the attempt to bring back normalcy in health as quickly as possible, a sufficiently large number of drugs, vaccines, and therapies are in development. Vaccines, however, take at-least a year and half to be safely and effectively applied; the short-cycle development of drugs and other therapies seem to take a successful lead by repurposing existing drugs, and known therapies. A big question for the medical domain, given the current urgency, is “which potential drugs or therapies are most promising to focus on?”.
Study results of these drugs and therapies have been reported in numerous scientific articles. In addition, a lot of useful information about experiences and early indications circulates the web, like blogs, social-media, and similar others that are openly available articles on the web. For example, here are two questions that may be answered by mining openly available information on the web: “We see and ncreased use of anti-malaria drugs in less effected countries. Is this indicative of a potential beneficial effect of those drugs?” or “Many people who have severe progression of the disease are posting messages about experiences of loss of taste and smell at the onset of the disease. Is loss of taste and smell indicative of disease progression?”. A collection of these articles after appropriate authentication of the sources may become knowledge sources for various purposes, especially for early hypothesis. Rapidly, mining these sources and engineering their knowledge may render the progressive outlook of potential drugs, vaccines and therapies in terms of its efficacy and safety. Our COVID-RAPID-MINER is a textmining tool with a dashboard that helps researchers to form hypothesis that are most promising for further study. Our approach will consist of knowledge discovery, smart trend analysis and intuitive presentation to assist researchers with intelligent reasoning, judgement and decision making.
Our specific objective for the hackathon is to develop a prototype tool, based on existing open-source components, capable of knowledge engineering the SARS-CoV-2 literature as well as less formal information scattered on the web and summarize opinions on the drugs, vaccines, and therapies and then visually present the results based on a pre-selected features for (known) safety and efficacy on humans. Our long-term motivation to this specific objective is to help stakeholders (e.g. health care sector) effectively plan, get subsidy on their needed actions (e.g. disease surveillance), and even get vigilantly prepared for the much anticipated second wave of COVID-19.