Methodology on the similarity analysis on doctoral theses elaborated within the Doctoral School of Mathematics and Computer Science (SDMI)


No. 8/25.01.2023

1. Preamble

Art. 1

This Methodology is based on the following document, supplemented by regulations specific to the fields of Mathematics and Computer Science:

  • Appendix 1 to the Regulations for the organization and conduct of doctoral studies at Babes-Bolyai University (approved by the BBU Senate Decision no. 483/13.01.2020 with subsequent amendments and additions made by the BBU Senate Decision no.33/14.03.2022) on the Procedure for carrying out the similarity analysis on the doctoral thesis approved by HCA no. 82/4.01.2022.

2. General provisions

Art. 2

  1. The purpose of drawing up the Similarity Report is to implement the principles of the Code of Doctoral Studies and the Regulations for the organization and unfolding of doctoral studies at the BBU approved by Senate Resolution no. 483/2020 with subsequent amendments and completion, as well as of the Code of Ethics and Professional Deontology of the Babes-Bolyai University concerning the observance of standards of ethics and academic integrity and the prevention and sanctioning of plagiarism.
  2. The standards of ethics and academic integrity and the prevention and sanctioning of plagiarism are regulated in Annex I of the SDMI Regulations, representing the Guidelines for the preparation of doctoral theses in the fields of Mathematics and Computer Science, updated and approved at the SDMI Council meeting on 25.01.2023 and published on the SDMI website.

Art. 3

  1. To check the doctoral theses for similarity, the analysis will be conducted using the Turnitin plagiarism detection system.
  2. The creation and the analysis of the degree of similarity of doctoral theses will be carried out before the pre-defense of the doctoral thesis and after the submission (to the SDMI secretary) of the thesis (in the final version, with the final title) by the Ph.D. students in electronic format, accompanied by the request for the start of the procedures of similarity analysis and evaluation of the thesis by the supervising committee.

3. Creating the Similarity Report

Art. 4

  1. The creation of the Similarity Report will be carried out at the Doctoral School of Mathematics and Computer Science, the process being supervised by the Council of the Doctoral School.
  2. The doctoral student, whose thesis is subject to the similarity analysis, and the SDMI secretary, under the supervision of the doctoral supervisor, will be in charge of generating the similarity report, taking into account the specific field in which the doctoral thesis is written (Mathematics or Computer Science).

Art. 5

  1. The list of bibliographical references and the original articles of the thesis author (self-citations) will be excluded from the Similarity Report to be submitted for analysis. Additional settings (e.g., setting a threshold to ignore texts below that length) will not be used.
  2. For non-English Ph.D. theses, a report on contextual similarities with English texts will also be generated. For these cases, an additional anti-plagiarism software setting will be used to generate the supplementary similarity report, which will be communicated by the SDMI secretary.

4. Similarity ratio analysis

Art. 6

  1. The Similarity Report results in a similarity coefficient (which determines the volume of borrowed text in the percentage found by the program in other documents) and a list of documents identified as possible sources.
  2. The doctoral supervisor will have access to the Turnitin software so that he/she can check directly at source the texts reported by the software as similarities, and not only their source.

Art. 7

  1. The doctoral supervisor will have to examine the Similarity Report in terms of the appearance of unauthorized takeovers in the thesis.
  2. In the SDMI, a percentage of 30% (alert threshold) for the Similarity Coefficient is used for the similarity analysis of doctoral theses. Exceeding the alert threshold does not necessarily indicate plagiarism, but further steps are required to explain the level of similarity.
  3. Even if the value of the similarity coefficient does not exceed the threshold set out in paragraph 2, the content of the Similarity Report must be checked for possible plagiarized parts which may affect the originality of the doctoral thesis.
  4. The doctoral supervisor must check at least the following aspects:
    • whether the work contains large portions of the text identified by the system as similar;
    • whether there are too many potential takeovers from a single source;
    • whether the paper contains editorial features that indicate the presence of unauthorized takeovers, which may be indicators of plagiarism.

Art. 8

  1. Following the examination of the Similarity Report, the Ph.D. supervisor will fill in the form called Resolution.
  2. If the similarity coefficient exceeds the alert threshold established according to Art. 7, paragraph (2), and the Resolution is “the work does not contain unauthorized takeovers”, the Ph.D. supervisor must explain the reasons for this resolution under the field Remarks.
  3. The Resolution must contain, under the Remarks, explanations/considerations by the Ph.D. supervisor of the similarity coefficient regardless of its value.
  4. The Ph.D. supervisor will not admit the work for the Ph.D. thesis if the evaluation of the Similarity Report shows that:
    • the takeovers that have been detected in the work are not justified and bear signs of plagiarism, and,
    • there are deliberate changes to the text in the paper, which indicate an attempt to disguise the unauthorized passages.

Art. 9

  1. If the Ph.D. supervisor accepts the Ph.D. thesis for the purpose of its defense, the Similarity Report must also be evaluated and assumed by the members of the supervising and defense committees, according to the decision of the Rector of the Babeș-Bolyai University of 618/17.01.2023.
  2. In case the supervisor and/or the members of the supervising committee make recommendations and suggestions to modify the Ph.D. thesis (in the Resolution and/or in the Protocol for the pre-defense of the thesis), then the whole similarity analysis procedure will be restarted and the thesis (in the revised form according to the given suggestions) will be re-entered in the similarity analysis software.
  3. The protocol for the pre-defense of the Ph.D. thesis will contain considerations of the committee members regardless of the similarity coefficient obtained from the Ph.D. thesis analysis.

5. Final provisions

Art. 10

  1. This methodology was discussed and approved at the SDMI Council meeting on 25.01.2023 and shall enter into force from the date of approval for theses to be filed for pre-defense after this date.
  2. Once approved, the methodology shall be published on the SDMI website.

Cluj-Napoca, 25th of January 2023
Prof. Gabriela Czibula, Ph.D.