Reading similarity reports from Urkund

WISEflow uses Urkund as plagiarism service. The text comparison (percentage) will be visible to external and internal assessors per default on the flow overview. The percentage appears in the participant list by each name and also in the assessor tool.

By clicking the percentage a new pane will open and the report is displayed. The report contains different information regarding the submitted paper.

Analysis Overview

When opening the report, the first thing you are greeted by is the analysis overview.

1. These squares represent the paper that has been analysed by Urkund. This provides an overview over which parts and pages of the paper has findings on it. Every coloured box here can be clicked to show more details about the findings.
   1. The arrow-circle is for resetting the report, in case you remove sources from the analysis or exclude different matching texts.
   2. The arrow pointing down lets you download a PDF-version of the entire report, so it can be viewed offline.
   3. The question mark will open a tour of the Urkund reports based on your current location in the report.
2. Findings provides you with a quick overview of how many matching text sections and warnings there is on the report. Clicking on "Matching Text" or "Warnings" will send you to the Findings tab, while clicking "View the entire document" will send you to the Entire Document tab.
3. Similarity shows you the submitted documents similarity percentage and the average percentage across all reports sent by your institution.
4. Submission details provides further information about the report, such as access to the original paper, when it was submitted, word count etc.

By clicking on either "Matching Text", "Warnings" or "View the entire document" you will gain access to a more detailed overview of the different parts of the report. This includes 3 tabs that you can switch between for the report.

Findings

Under the Findings tab, you will have access to each matching text and warning in a side-by-side view with the submitted document on the left and the source of the match on the right. This way you can focus on going through the matches without having to scroll through the whole document.

Matching text are blocks of text where a match was found in a source.

Warnings are indications of suspicious items in the submitted document. It could bad PDF-formatting, suspiciously long words to downplay the word count or use of special characters not noticeable by the reader.

1. Here you can switch between the finding types, Matching Text and Warnings.
2. Here you will find the option to highlight quotes, brackets and detailed text differences in the text. If the "Detailed text differences" is turned on, the differences is highlighted in the matching text. Words highlighted with yellow means that they have been substituted and red means they have been removed.
3. The left side shows the section of the submitted document that has been flagged for having similarities with another document. At the top of this box you have the option to exclude this section from the analysis, if the match is false positive.
4. The right side shows the text in the matching source. The percentage at the top show the similarity of text for this specific block of text. At the bottom you can see information about the source and the icon shows what kind of source you are dealing with. W meaning website, SA is a student assignment and J is content from a journal.

On the findings tab you can navigate through the different matches by using the left and right arrow keys or by clicking the next and previous buttons.

Sources

Website

The Internet contains billions of pages with various content; everything from material published by universities and colleges to government agency material, press articles, books, reference works and much more. There are also specific cheat sites online with ready-produced material.

A quantity of the material available on the Internet is only accessible through password-protected systems; hence, it cannot be located with the aid of an ordinary search engine. There is also a plethora of search engines, each of which with its own coverage. Furthermore, there is a huge amount of material on the Internet that cannot be found using regular search engines. URKUND manages to find sources through all mentioned above.

Journal

Published material consists of hundreds of millions of books, reference works, scientific articles, and so forth. Some material is accessible electronically via specific databases, whilst other material can only be found in the printed format. URKUND has formed strategic partnerships with a number of leading information providers.

School Assignment

Plagiarism is, of course, also prevalent between students themselves. Examples of this are that a student may plagiarise material from someone who has previously studied the same courses, at the same or at another school, or that through means of close collaboration, two students may copy each other’s’ work prior to a submission. As a rule, such student material is generally not published. Consequently, it cannot be searched for on the Internet or in published material, but Urkund stores all of these assignments.

The sources tab gives you access to all the sources where similarities with the submitted document have been found.

1. These icons tell you what type of source is referred to. W = website, SA = student assignments (documents previously submitted to Urkund stored in our archives) and J = journal, referring to published books and other physical medias that have been scanned in.
2. This is the list of identified sources where a match has been found. It shows the similarity percentage, where it was found, the number of matching parts and where these similarties are located in the source. There is also an option to remove the source from the analysis if there is a need.
3. If you find a source on the internet which Urkund has not detected during its search you can use this option to add it. Do keep in mind that Urkund does not have access to sources on the internet that are locked behind paywalls/passwords etc.

Entire Document

Under the Entire Document tab you can see the document in its entirety with all of the findings highlighted in the text.

While viewing the entire document "Matching text" and "Warnings" are highlighted both with colour in the text and with corresponding icons to the right. Clicking the icons will show you the matching text from the source and information similar to those under the Findings tab.

Similar to the Findings tab you the option to highlight quotes, brackets and detailed text differences in the text. If the "Detailed text differences" is turned on, the differences is highlighted in the matching text when you have clicked on an icon. Words highlighted with yellow means that they have been substituted and red means they have been removed.

Conflict copies:

Sometimes a student might be in need of adding or handing in their assignment again. If a hand-in has been administratively pulled back, it might already have been scanned and saved by Urkund. When the participant hand-in their assignment again, it would then be reported up against the previous version of itself.

In many cases this can lead to high similarity reports, where it has flagged almost entire documents as being 100% similar to another report. In cases such as these, Urkund is smart enough to see, that the paper giving a high similarity is from the same exam and participant. It then adds this as a "Conflict Copy".

With this option from the main page in Urkund, you can easily turn off the conflicting copy, so more legit source similarities can be shown in the areas that the conflicting copy was taking up unnecessarily. The report will then show its true percentage and change to show the other sources instead.