Earlier this year, a group of refugees showed up in a supermarket in Groß-Gerau, a southwestern German district not too far from Frankfurt.

They filled up their shopping cart and said they wouldn't be paying for anything, informing a surprised cashier that Angela Merkel, the country's chancellor, would be picking up the tab. Following a standoff, the mayor of the town showed up and paid for the refugees' big shop.

This would be a remarkable anecdote of how asylum seekers are taking advantage of Germany's "Wilkommenskultur" — if it weren't for one detail.

It never happened.

The rumor is one of almost 300 to have been debunked on Hoaxmap, a simple project set up this month by two Leipzig-based Germans in their thirties, Karolin Schwarz and Lutz Helm. The two decided to launch the project as they saw that last summer "the numbers of refugees were rising, and so were the rumors about refugees," they told me via email.

Schwarz and Helm were concerned that while rumors were being shared widely over social media, corrections weren't receiving as much attention. They hoped Hoaxmap could serve as a "database to look them up."

Other rumors debunked on the site include stories about refugees abducting and eating dogs and the purchase of 140 euro shoes on the taxpayer's dime.

Many involve darker issues such as murder and rape. The rumors have fallen on fertile grounds following the indisputably real assaults and robberies in Cologne over New Year's Eve, which were initially mishandled by local authorities.

Real crimes can serve as inspiration for fake ones, however. The abduction of a 13-year-old Russian-German girl by a group of migrants led to an international outrage - until it was discovered the girl had actually absconded to her boyfriend's house.

In this context, it is no surprise that Hoaxmap has attracted plenty of attention. The site launched three weeks ago but has already been profiled on Al Jazeera, CBCDeutsche Welle and Forbes.

Schwarz and Helm launched the site with an initial 177 refuted rumors on February 8th. Since launching they have added over 100 rumors received via email or Twitter — and they have a growing backlog to verify.

Their method is not flawless; with no resources besides their free time, they have to rely on reports rather than first-hand investigations (Schwarz is a consultant, Helm a software developer).

"We accept statements from police or administration spokespersons, news articles from well-established media organizations, and also statements by involved parties, for example statements from spokespersons of the stores that have supposedly closed due to thefts or attacks," Helm says.

The two also rely on other online debunkers like mimikama.at, an Austrian debunking website (the map covers Austria and Switzerland besides Germany).

In an interesting sign of unwanted consequences and the media's responsibility in spreading rumors, the Hoaxmap creators have noticed that "false reports in news articles are often initiated by calls for witnesses in police reports after a supposed crime."

While a majority of the rumors on the map seem to be ones that would be spread by detractors of the Wilkommenskultur, some go the other way.

One of these cases involves a rumor about a Syrian refugee who had supposedly died because he was waiting for too long outside the Berlin LaGeSo (where asylum seekers first file their application for refugee status). The death never happened.

With the horrific conflict in Syria ongoing and the number of asylum seekers in Germany in the hundreds of thousands, this issue is unlikely to lose its relevance soon.

I asked the Hoaxmap duo how long they intend to have the project running. Helm said "we have no precise plan on how long we will continue this project. If the numbers of rumors are dropping there might no longer be a need for this project. On the other hand, a database of rumors might be permanently useful."

As Hoaxmap's database grows, it may be interesting to use their structured findings to learn more about the overall trends in rumors about refugees. It may also be worth applying instruments like RumorLens to evaluate how far certain rumors go on social media and which debunks seem most effective.

Helm is interested, but appropriately cautious. "Our collection is not quite complete, and it probably never will be because rumors are a very vague thing. It is even too early to make any assumptions about the local distribution of the rumors." That said, his hope is that "people will use this data in ways that we are not aware of at the moment."

"The other" has historically been a rich source of rumors born of unfamiliarity and prejudice. But large-scale migration is not the only phenomenon that could be well-suited to a topic-specific Hoaxmap-like project in other countries: Health scares like the Zika virus seem like an obvious fit.

In the long run, collecting rumors and coupling them with their debunks in structured thematic databases may help us understand how to best prepare ourselves to combat rumors in the future.