Analyzing a COVID-19 Article

This articles provides a couple of snapshots of a recent experiment when analysing a coronvirus article using Semantic Brain technology and practices.

Step 1: Data Ingestion and Pre-processing

• Ingest article – https://www.livescience.com/is-coronavirus-deadly.html
• Convert to plain text
• Perform initial natural language processing

Step 2: Enhanced Processing and Data Extraction

Further natural language processing combining rules and neural networks extracts

Age Group	Fatality Rate
10 to 39	0.20%
40 to 49	0.40%
50 to 59	1.30%
60 to 69	3.60%
70 to 79	8%
> 80	14.80%

Step 3: Question Answering

User asks the following question “What is the COVID-19 death rate by age group?”. The result

Conclusion

The above result was achieved by training the system to recognize infectious diseases, health impacts and age group recognition. No clinical subject matter expert was consulted.

With further clinical SME input and training it should be possible to answer complex questions relating to healthcare and medicine.