Crowdsourcing for Science : Team AKA
State the Challenge Number and Project Name: Team AKA
Our project aims to improve a COVID-19 fact checking website called "check-the-fact.com". The current website only allows certified users to fact check information that is mainly in German, which narrows the user base for this global crowdsourcing activity.
Our team is called AKA, Abdallah Itani, Kevin Adea, and Aleksander Michonski. We are Computer Science Graduate students at the university of Geneva.
Describe what is the need of this project?
We wanted to improve this fact checking website since we are currently facing a global issue that requires accurate information and proper awareness.
Currently the website only allows certified users to fact check facts that are in German and if potentially more data comes in in different languages too, the information in the manual voting section is mixed and is prone for error. And for the global event we are living in, a website like this should allow accurate fact checking in the languages the voter chooses to prevent misclassification.
What is the evidence? Also who can you interview? What can you find out? What experiment can you run?
We think everyone has at least witnessed fake news once during this global event, especially on texting apps such as Whatsapp, and news outlets reporting fake news.
We can interview users of such social media apps and ask them how often they encounter fake news, and how certain were they from the news they receive.
An experiment that can be made is providing the people we interview with a set of facts on COVID-19, and ask them if they can correctly identify which is fake or not, based on the information they have been ingesting from the internet.
And regarding the website, the language barrier problem can be seen in the "Vote Fakenews" section of the website. We can ask people to fact check something in the website aswell to confirm that the database is mostly in 1 language, the website will most probably won't be able to check for inputs other than German.
Start documenting your thoughts and ideas
- Collect data from people on the spread of fake news to confirm that it is indeed a problem that needs tackling.
- Contact the developers behind the fact checking website, and ask them whether they can provide us the source code to work on, or to help them with providing the data needed to improve the fact checking.
What is The Big Idea? What is the value proposition?
The problem is that although this website that was created has the correct idea of implementation, it is currently not producing much results because of two issues: No database to start off with (so unless many people are helping the machine won't generate results), and there is a language barrier that disables many people from helping (currently mosty in German, and when more data comes in, it looks like the languages are mixed).
Our big idea is to produce a good database that the website can start with, and for any facts that are then asked to be proven through votes, we want to create a filter that allows people to be crowd sourced only in their language.
What is the mechanism of beneficial change?
Through this approach, we will be able to have a website that functions, and for anyone who actually wants to prove whether a certain piece of news is fake or not, our website is there to help.
What are the key metrics?
We will be able to measure the impact through the amount of usage in the website. At the moment, we do not have access to the website's statistics as we have not yet received a reply from the creators of the website. However, we believe that if through this project we still do not receive any contact from the creators, we may be able to estimate the impact through how many facts are being asked to be proved and voted on.
Who is most likely to be supportive?
Those with the greatest interest in this are the community. This website was made specifically with the community in mind, so that they aren't victims to false information.
Key foes? Who is most likely to oppose?
Those who are spreading the false information in the first place. For whatever reasons they are doing so, they would be opposed to this website as it would mean a less effective spread of false information.
This has made as also think that these people who are spreading false information can also use our website (Vote for the fake news on purpose), so that their fake news can be more "credible". However that is an issue that we will not be focusing on in this project.
What is the user experience?
A user can input a message, file, or link in the website and they will find out if it is fake or not. However if there is not enough information regarding this fact, it will go to a voting poll on the website and people who would like to help can vote for the correct fact.
Registered users can also report news themselves provided they include a credible source
Who has to do what. to make it happen?
The website must first get a good database to start with. This will allow the website to be more useful and as such get more users. Having more users also means more potential crowdsourcing since the website gets more traffic.
Even if the website separates the facts by language by any mean (google translate), getting a good database for each language is a must, and it produces more accuracy.
Who are the key partners to execute? Key partners to help others evaluate your value preposition?
Projects related in state of the art (links in literature), websites with the basic common questions about the corona which sum all them up and provide reliable answers.
Either teams or individuals who are currently working on building/ or have built a fact checking website like https://www.check-the-fact.com/ and require more data.
What are the precipitating events?
In the current abrupt turnout of events we need a fast and reliable source of information verification. In order to precipitate the process we propose to engage crowd aspect by sending links to fill out databases in different languages among our friends and social medias.
who else is in the field?
Physical and intellectual resources needed (besides financial resources)
What is needed is a database of good size containing facts and myths (sentences, links to articles and magazines, websites), and not only in a single language, but multiple.
And it is required from the community to be aware that their daily dose of information from the Internet is usually a mess of facts and myths and there is no way for them to be sure unless they dig deep, most people do not and just simply choose to accept what they see online or what their friends have forwarded to them.
Next steps? Pilots?
Writing a demo that would show working project from a big picture perspective. Send it to the relevant working groups to gather their feedback.
How might this go wrong? How might the problem evolve? What are the legal, cultural and other impediments?
1) Database is built up by people. They do not specific knowledge neither they are checked for trustworthiness of their answers. People may answer wrong on purpose. How to supervise it? It is a basic problem of many projects based crowdsourcing contribution.
2) Fallibility of AI algorithm deciding, sorting and trying to understand if the question exists in the database.
How will i promote adoption?
By sharing the link for the project in social media, friends, class, Get in touch with non-profit organisations related to our project theme i.e. Covid-19