Urkund is a fully-automatic machine learning text-recognition system made for detecting, preventing and handling plagiarism, no matter which language you are writing in.
There are many reasons why Urkund is trusted by thousands of institutions in over 70 countries worldwide. Being a reliable system with cutting-edge machine learning algorithms is one of them.
EASY TO USE
Urkund was created by teachers for teachers, to make their life easier. That is the core of pour business and part of every developing process.
Urkund delivers accurate reports quickly, even in hectic exam times. Our aim is to save our users time while doing the job of detecting plagiarism in the best possible way.
GDPR Compliant with privacy by design and an uptime of 99.9%. You can trust Urkund to be your partner in academic integrity.
HOW CHECKING A DOCUMENT WORKS IN URKUND?
The Urkund process starts when you submit a document to the system. This can either be via the schools' existing learning platform or as simple as attaching it and sending it in an email. The system supports a wide variety of the most popular file formats, your standard word processors, as as archives of several documents when emailed. The text is then extracted from the document and the processing begins no matter what language the text was written in.
RELEVANT SOURCES ARE RETRIEVED
The text is analysed and we begin to check for possible candidates for text similarity across our three sources areas. Urkund singles out potential matching sources from our archives containing sources from the wider Internet, academically published material, and previously submitted student documents. Candidates for text similarity are studied in depth for similarities and any results are saved.
MACHINE LEARNING POWERED ANALYSIS
When a document starts to display similarities to other sources, the basis for a report beings to take shape. We not only record the degree of the matches but take into account any usages of paraphrasing and synonyms as well as other form of substitutions. Our algorithms work regardless of language and are even able to detect similarities across a growing number of languages.
REPORT CREATION AND DELIVERY
When the process is finished an analysis overview is generated and delivered to the teacher, either via the used learning platform, by email or through the web inbox. The results of the analysis are presented in a comprehensible report so that the teacher can make the final judgement if any plagiarism has occurred or not.
OUR THREE SOURCES AREAS
In order to prove that a piece of text has been plagiarised, finding the sources is the greatest challenge. The original content that is plagiarised can be found in three different source areas: The Internet, Published Materials and Student Materials. The common denominator in all three categories is that they contain an immense and ever-increasing amount of source references, which makes it close to impossible for teachers to find it unaided.