Updated: 03/02/2020

An outbreak of coronavirus disease 2019 (COVID-19) caused by the 2019 novel coronavirus (SARS-CoV-2) began in Wuhan, Hubei Province, China, in December 2019, and has spread throughout China and to 31 other countries and territories, including the United States (1). As of February 23, 2020, there are 76,936 reported cases in mainland China and 1,875 cases in locations outside mainland China. There have been 2,462 associated deaths worldwide; no deaths are reported in the United States. Fourteen cases have been diagnosed in the United States, and an additional 39 cases have occurred among repatriated persons from high-risk settings, for a current total of 53 cases within the United States. The mortality rate worldwide is 1 in 32 cases, which is slightly more than 3% – very similar to “common” illnesses like influenza.

Social media analytics has the power to illuminate insights into this public health crisis and assist in maintaining situational awareness on breaking events. One example of social media playing a role in increasing situational awareness is the creation and distribution of a map produced by researchers at Johns Hopkins University. The dashboard tracks and visualizes reports about the coronavirus outbreak using data from the Centers for Disease Control and Prevention, the World Health Organization, China’s CDC.

Monitoring 2019 novel coronavirus (2019-nCoV) on social media with DigitalStakeout Scout

As the number of 2019 novel coronavirus or 2019-nCoV cases rises, it’s understandable social media posts about the illness will rapidly increase the share of conversation online. This will result in a lot of false, inaccurate information to spread while creating confusion about what is accurate and reliable information, which has a synergistic effect on the footprint of the outbreak. Overreaction breeds more overreaction, and that’s where a tool to cut through the noise will help with a more objective analysis of the situation.

In this post, here are some recommendations for analysts on how to start the process of monitoring coronavirus in DigitalStakeout Scout.

Monitor the following coronavirus hashtags

  • #coronavirus
  • #covid19
  • #coronaoutbreak
  • #coronavirusupdate
  • #covid

Maintain a point of context, coronavirus vs flu

According to the CDC, there have been over 25 million cases of the flu this season and up to 25,000 deaths. The current flu season is not being reported as severe. When creating a monitor in DigitalStakeout Scout, create a monitor for “coronavirus” AND a monitor for “flu” to benchmark against each other. Using flu as a barometer will enable you to review information about coronavirus with an objective approach. Using Scout Analytics, compare the total volume, number of participants, and the general sentiment of the two topics.

Monitor your industry & for breaking events

Use Scout to identify real-time news about your vendors, your industry, and track dynamics that could impact your supply chain and production. DigitalStakeout Scout automated discovery feature will tag high-risk posts that apply to numerous risk categories that are a sign of physical and operational hazards.

  • Create a Keyword monitor and follow trusted news sources such as @AP and @Reuters.
  • Create a Keyword monitor to detect #breaking AND check the verified account option.
  • Create the necessary Brand monitors to track mentions about core suppliers, vendors, and partners.
  • Create a wordlist and correlate critical vendors & suppliers against your coronavirus feed.

Monitor by location, leverage the Public Health word list

Location-based social media analytics provide the opportunity to detect public health conversations at target locations and enable you to take action to mitigate risk to the people, property, and the place. Use Location Monitors to define vital geographic areas and monitor multiple locations simultaneously. To isolate conversations about the coronavirus, use the Public Health word list to filter results.

Filter out unverified accounts, retweets, and bots.

Scout enables you to perform a unified analysis on all of your monitor data or data from a single monitor. You can search your real-time and historical monitor data by full-text search by an entity and by tags visually without having to learn a technical query language. To reduce the risk of reviewing false information, filter your Twitter results by Account Status field to filter “verified” and the Action field as “post” and Generator type to “social_mgt,” “web” and “mobile.” Filter your results to Verified Twitter accounts, exclude retweets and exclude application and bot-generated content. Apply this filter to any Monitor data you review or analytics that you create.

Monitor feeds and updates from reputable health sites.

The RSS Monitor regularly monitors and centralizes all the health site feeds you want to track in one place. All that is required to use this monitor is a valid URL to an RSS or ATOM feed. RSS will regularly poll the sites for updates and unify the posts with all your other Scout feeds.

CDC: https://tools.cdc.gov/api/v2/resources/media/403372.rss

WHO: https://www.who.int/feeds/entity/csr/don/en/rss.xml

TRAVEL.STATE.GOV: https://travel.state.gov/_res/rss/TAsTWs.xml

Certain websites that do not have RSS feeds do have email subscriptions and do not update social media with detailed information. In this situation, configure and Email Monitor to create a virtual inbox to receive email from subscribed sources. Once you create your virtual mailbox, use the email address provided to subscribe to relevant sources of information.

National Institute of Allergy and Infectious Diseases (NIAID): https://www.niaid.nih.gov/global/email-updates

Watch for highly repeated non-unique content and where it links to.

In any crisis, fake news and malicious content a part of the equation. It is very easy to fall victim to fraudulent content unwittingly. To combat this threat, Scout detects high-risk content that you would not want to click or include in your analysis.

  • Scout stores a content hash of each post discovered. Watch for high counts of “content hash” in the derived fields as this would be an example of the exact content being shared by numerous accounts. Be wary of this content sourcing from un-verified accounts.
  • Scout checks the reputation of the domain and IP address hosting the content is checked for security concerns in real-time. Secondly, the content host information is attributed and geo-tagged. Do not click on any content marked as “cyber exposure” and “compromised.”

Combined with the recommendation point #4, use this tactic to protect yourself from engaging and sharing fraudulent content.

We hope this post gives you ideas on quickly standing up monitoring and cutting out the noise.

If you need any assistance with this matter, please contact us and a member of the DigitalStakeout team will follow-up with you immediately. DigitalStakeout can provide your organization with a quick ramp monitoring solution to monitor coronavirus outbreak developments specific to your situational awareness needs. We can have your organization fully deployed from this template in less than a day.