How COVID-19 Check predicts which “F” you are?
As far as you may know, Got It engineers are volunteering to build the COVID-19 Check application to join hands with the Vietnamese community’s efforts in reversing the spread of Coronavirus. After the public beta was released to test and validate the features of the system, our engineers received many valuable comments from the community. Since then, they have been working hard and actively with the support from many partners to deploy COVID-19 Check and serve the community as soon as possible.
While waiting for the official version, let’s find out how COVID-19 Check works in detail to know the way COVID-19 Check predicts which “F” you are with our below article. As this is a community project, we also hope you can join hands with the Got It team to make the app beneficial to users and Vietnamese anti-epidemic organizations. Many a little makes a mickle, so there’s no need to be afraid of Coronavirus!
Common forms of Coronavirus’s infection in Vietnam and ways to prevent the spread of the virus
Until now, all cases of Coronavirus infection in Vietnam are the result of close contact between infected people moving from epidemic areas and other people when they shared public transport. Since March 6, 2020, most cases of Coronavirus are people who come from foreign countries to Vietnam, are “imported” by air and then spread the virus to others in the country as described in the infection map of Tuoi Tre Newspaper below.
In the first 16 virus infections, Vietnam has done a great job of stopping the spread of disease for a long time through serious quarantine, regardless of which direct level people communicate with each other. Coronaviruses will spread at a rapid speed if the government doesn’t control and promptly isolate infected cases, such as the case of patient number 34.
To make the quarantine effective, the Ministry of Health in Vietnam has a classification system from F0 to F5 as illustrated by VietnamPlus.
F0 —Person who is identified to be infected with SARS-CoV-2
F1 — Person who is suspected to be infected or exposed to F0
F2 — Person who is exposed to F1
F3 — Person who is exposed to F2
F4/5 — Person who is exposed to F3/4.
The quarantine of infected, suspected or potentially infected people is a vital action to prevent the spread of Coronavirus to the community. Because of the extremely rapid spread of Coronavirus, not only in Vietnam but also in many areas of the world, the quarantine in different levels is enforced: Italy closed all borders, Wuhan (China) blocked the whole city, San Francisco (USA) asked people not to go on the streets. All actions are aimed to reduce the dramatic rise of patients (flatten the curve) and to avoid the collapse of the health system due to overload.
Quick identification and classification of individuals from F0 to F5 will be helpful to assist the anti-epidemic organizations in implementing the quarantine appropriately as well as letting the individuals know how likely they can be infected in advance to make some preparation.
Why was the Graph selected for COVID-19 Check?
As mentioned above, the key to finding F1 to F5 quickly is to build related contact points, which includes 3 types:
- Contact related to close interpersonal communication: It is defined as two or more individuals contacting in close distance (≤ 2m) and at risk of Coronavirus infection if one of those individuals tests positive. Some examples of close connections include shaking hands, talking, meeting, etc.
- Contact related to an individual’s journey: As we have seen in the media about the VN054 flight, a lot of people in this flight are infected with Coronavirus. And recently, Vietnam’s Ministry of Health has made a lot of urgent announcements to find passengers on specific flights because not until later did someone test positive for Coronavirus. Some examples of such connections are people taking the same flight, bus, or yacht.
- Contact related to places individuals are present: There are several cases where individuals become infected with Coronavirus when they appear in an epidemic area, such as patients 17 and 18. Another example is the patients 22, 23 and 35, who all went to the “Dien May Xanh” Supermarket.
To make it clearer, we can look at the case of patient number 17, who has all three types of contact points:
- Contact related to close interpersonal communication: Patients numbers 19, 20 and 47.
- Contact related to an individual’s journey: Flight VN054.
- Contact related to places individuals are present: Epidemic areas in Europe: Milan ((Lombardy), London, Paris. Besides, she also has one more contact related to places, which is the address 125 Truc Bach, Hanoi.
These contact points can be demonstrated easily by Graph consisting of node and relationship as follows:
As you can see, this is a part of the graph that illustrates patients on flight VN054 and infected people who contact those passengers afterwards.
This is the graph of F0. If there is any direct connection of other individuals to this F0 graph, those individuals will immediately become F1, similar to F2, F3, etc. VN054 was confirmed to have a person infected with Coronavirus so all other passengers on this flight could be considered as F1. Similar to what happened at “Hong Ngoc” hospital because the patient number 17 was here, everyone appearing at that time could be identified as F1.
The graph is one of the best methods to demonstrate these relationships, and when networks are established, finding, predicting and localizing, etc will be carried out very quickly with algorithms of graph theory.
In terms of technology, there are databases created specifically for storing graph data, such as Neo4j that can handle extremely large graphs with billions of relationships. This is also the basis that helps Got It engineers choose the platform to build COVID-19 Check.
How COVID-19 Check works
The main objective of COVID-19 Check is to build a large-scale graph database to archive contact points between individuals within 14 days (14 days is the common incubation period of Coronavirus, after 14 days, the contacts are meaningless).
This graph database is built based on the combination of different data sources:
- User-provided data
- Official epidemiological data from the Coronavirus Anti-Epidemic Steering Committee
- Social media data from participating partners.
After the network has been established within 14 days, if there are any changes on the network, the calculation for the related nodes will be done immediately. Here are some specific examples:
- When an individual (node) is confirmed to be infected with Coronavirus (F0), F1, F2, F3, etc will be immediately found.
- When a flight is confirmed to have passengers infected with Coronavirus (F0), all passengers on that flight (F1) and their contacts (F2), etc. are also found promptly.
- When a place is confirmed to have people infected with Coronavirus (F0), all people who arrive at the same time with the infected patient are immediately found.
Once in the network, an individual always knows how he relates to infected people (which F he is) based on the relationships he has with individuals, places, or journeys that are already archived in the network. Therefore, the more individuals involved in providing exact data, the more accurate the calculation results of COVID-19 Check are.
How COVID-19 Check can help you and other anti-epidemic actions
COVID-19 Check is designed to assist in providing information for different parties when the whole country participates in reversing Coronavirus:
- For users: If everyone regularly updates their interpersonal contacts, places and journeys within the past 14 days, they will always know their risk of infection at any level from F0 to F5 +. Therefore, if the information is updated daily, users are the first side to benefit from that database to prepare for different situations.
- For government’s anti-epidemic departments: This network will help the government to localize, predict and prepare for quarantine or treatment of potentially infected individuals as well as disinfection of unsafe areas to prevent the rapid spread of the virus to the community.
- For researchers: The COVID-19 Check’s network will provide information that can be applied in analysis and the use of AI algorithms to assist in other anti-epidemic activities.
How to use COVID-19 Check
As for users, COVID-19 Check works in a very simple way:
1. Users create accounts by inputting their mobile phone number into the system. Mobile phone number is the only identification that users provide to authenticate and avoid spam.
2. Users use the feature “Check-In” to enter data in the past 14 days, including:
- close contact with others
- journeys using public transport
- locations they have been present.
The first turn users inputting data may take time as the system requires data in the past 14 days; however, afterwards, users can easily update their data every day when an event happens.
3. Users check their risk of infection (which F they are). The system also automatically notify users if there is any change in their risk to help them prepare necessary plans.
How is the data privacy guaranteed by COVID-19 Check?
COVID-19 Check takes great care to protect users’ privacy as well as personal data and is designed based on the following principles:
- The system only requires a mobile phone number or an email to authenticate the user. If users use multiple mobile phone numbers, they can add these numbers to let the system know and manage contacts more accurately. Also, the system does not require users to provide any additional personal information, including Name and Address.
- Contacts are also created via mobile phone numbers or email identifiers.
- Contacts that last more than 14 days compared to the current day are automatically deleted from the system permanently.
Our next development plan and what can you do to join hands with COVID-19 Check?
COVID-19 Check is designed to handle huge amounts of data with hundreds of millions of nodes (equivalent to Vietnam’s population) and billions of relationships. The system is also built with an open plan so that it can be easily integrated with other data sources such as the huge social data of Kompa / Filum or data sources from other anti-epidemic parties. Thus, the network can maximize its usefulness for both users and organizations participating in the process of preventing Coronavirus.
In the first phase, Got It engineers already developed the backend and web frontend part, then tested it with very good results. However, the web frontend still has many limitations for users, especially in data entry because most of them must be entered manually.
In the upcoming time, the team will continue developing the mobile application to make the “Check-In” step become automated and create more utilities for users, thereby reducing the laziness of updating data due to too much manual work.
If you would like to join Got It to help reverse the Coronavirus pandemic, don’t hesitate to contact our team via email: covid19@got-it.ai. Got It will publish the Trello board soon to let everyone know the roadmap of this entire project.
______________________________
If you have any questions about COVID-19 Check, don’t hesitate to text us via our fanpage Got It Vietnam or the email address covid19@got-it.ai.
Follow Got It at:
- Fanpage: fb.com/gotitvietnam
- Website: www.got-it.ai /vn.got-it.ai
- Lever (job opportunities): bit.ly/gotit-hanoi
- Instagram: @lifeatgotit
- Email: vnteam@got-it.ai
