Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
Next revision Both sides next revision
en:interpretinganalysis [2016/01/21 15:15]
gustavo created
en:interpretinganalysis [2018/09/27 15:06]
gustavo
Line 1: Line 1:
- 
- 
 ==== Plagius - Interpreting the Results of the analyzes ==== ==== Plagius - Interpreting the Results of the analyzes ====
  
-{{  http://2.bp.blogspot.com/-XBcOF2OuIIw/UpP4s_Lbs8I/AAAAAAAAD0c/q2UhEYrj61w/s200/lookingatbook.png?nolink|}}+{{  https://2.bp.blogspot.com/-XBcOF2OuIIw/UpP4s_Lbs8I/AAAAAAAAD0c/q2UhEYrj61w/s200/lookingatbook.png|}}
  
-The Plagius is an excellent tool to assist in plagiarism detection, but it is still necessary to apply some criteria in its results.+The Plagius is an excellent tool to assist in [[http://www.plagius.com/en|plagiarism detection]], but it is still necessary to apply some criteria in its results.
  
 An automated system, such as Plagius, is not yet able to affirm plagiarism directly, only a manual scan can confirm the actual occurrence of plagiarism. An automated system, such as Plagius, is not yet able to affirm plagiarism directly, only a manual scan can confirm the actual occurrence of plagiarism.
Line 12: Line 10:
  
  
-=== How to interpret the test results ===+=== How to interpret the analysis results ===
  
 Despite Plagius brings diverse information in the analysis result, the best approach is still the manual review of the analyzed text: if there are many highlighted phrases together (regardless of color) next to each other means there is high chance of this section have been copied. Despite Plagius brings diverse information in the analysis result, the best approach is still the manual review of the analyzed text: if there are many highlighted phrases together (regardless of color) next to each other means there is high chance of this section have been copied.
Line 21: Line 19:
  
 When there is a suspicion, the highlighted section must be manually compared to the original source to ensure that is a case of plagiarism. When there is a suspicion, the highlighted section must be manually compared to the original source to ensure that is a case of plagiarism.
 +
 +[[https://checkingplagiarism.wordpress.com/2017/01/25/plagius-plagiarism-detector-2-4/|{{ :en:result-en-with-tips.png?direct&800 |Plagius Result Screen with tips}}]]
  
  
Line 26: Line 26:
 The percentage represents only a statistical indicator. Reports the percentage of the document sentences that were located on some website. It cannot be considered as plagiarism identification because not always that a phrase already exists on the Internet is plagiarism. A common example is the citations: if there are a high number of citations is higher the percentage of suspected plagiarism, regardless of whether the citations are properly referenced, so the percentage of suspicions will differ from the reality. The percentage represents only a statistical indicator. Reports the percentage of the document sentences that were located on some website. It cannot be considered as plagiarism identification because not always that a phrase already exists on the Internet is plagiarism. A common example is the citations: if there are a high number of citations is higher the percentage of suspected plagiarism, regardless of whether the citations are properly referenced, so the percentage of suspicions will differ from the reality.
  
-FIXME ** This page is not fully translated, yet** \\ // (Remove this paragraph once the translation is finished) //+[[https://checkingplagiarism.wordpress.com/2017/04/21/why-it-is-a-mistake-to-use-percentage-to-measure-plagiarism|Why it is a mistake to use percentage to measure plagiarism]] 
  
 === Variation between different analyzes for the same document === === Variation between different analyzes for the same document ===
-Every analysis of a document, the result will always vary because the search phrases are built randomly and search engines are also chosen randomly. The difference between the results does not compromise the result of the analysis.+In the analysis of a document, the result will always vary because the search phrases are built randomly and search engines are also chosen randomly. The difference between the results does not compromise the result of the analysis.
  
-=== When there is actually Plagiarism? === +=== When is there actually Plagiarism? === 
-Not based on any study but in my own experience, for any plagiarism claim is to be proved. Usually this is done showing the original text and where this text was inserted into the work / document under consideration, without proper references.+Not based on any study but in my own experience, to affirm plagiarism you need to prove it. Usually this is done showing the original text and where this text was inserted into the work/document under evaluation, without proper references.
  
-=== Colors of the test result === +=== Colors of the analysis result === 
-In the analysis results, the generated text is displayed in various colors, ranging between black and all shades between blue and vermelho.Em short, the closer the greater the amount of red encontraras occurrences on the Internet for that stretchThe test result there is also a "Legend" button that helps you remember what the colors. See the Figure below:+In the analysis results, the generated text is displayed in various colors, ranging between black and all shades between blue and redSumming up, the closer to the red color higher is the number of occurrences found on the Internet for that sentenceIn the test result there is also a "Legend" button that helps you remember what the colors means.
  
 === Other approaches === === Other approaches ===
-In the second tab of the test result, "frequently addresses" the Plagius lists the addresses found ordered by the frequency with which they were obtained. These addresses are strong candidates for a manual review, especially when the similarity percentage is significant.+In the second tab of the analysis result, "frequently addresses" the Plagius lists the addresses found ordered by the frequency which they were obtained. These addresses are strong candidates for a manual review, especially when the similarity percentage is significant.
  
-To assist this individual comparison, Plagius has a feature which helps in this task: Right click on any item in the list and choose the "Compare". This feature is best explained [[http://blog.plagius.com/2013/11/plagius-estabilizado.html|neste post]].+To assist this individual comparison, Plagius has a feature which helps in this task: Right click on any item in the list and choose "Compare". This feature will be best explained in a new Wiki topic.
  • en/interpretinganalysis.txt
  • Last modified: 2023/01/12 07:22
  • (external edit)