Skip to main content
v2026.11,986 entries · CC-BY 4.0

Unicheck: Reading Similarity Reports

Unicheck scans academic papers against web and institutional repositories, generating color-coded similarity reports with automated citation exclusions.

Written and maintained by CASRAI Editorial Board

Published · Last updated

On this page

Purpose of the section

Similarity screening during manuscript preparation and submission identifies textual overlap between a draft and published academic literature, open-access repositories, and student coursework archives. Unicheck (originally launched as Unplag by Phase One Karma in 2014, rebranded as Unicheck in 2017, and acquired by Turnitin on June 5, 2020) is a cloud-based similarity detection tool designed for integration within learning management systems such as Canvas and Moodle. In scholarly writing and pre-submission workflows, the similarity checking section allows authors, advisors, and editorial screeners to verify original phrasing, catch missing attributions, and distinguish legitimate academic citation from textual plagiarism.

Unlike journal-focused screening platforms like iThenticate that specialize in publisher cross-checks, Unicheck evaluates student theses, capstone dissertations, and course assignments directly within institutional portals. Its scanning engine inspects submitted texts in real time against online resources indexed by search engines (including Yahoo and Bing) and institutional databases. The primary purpose of interpreting a Unicheck similarity report is not to assign an automated verdict of misconduct, but to audit manuscript mechanics: confirming that quotation marks enclose direct excerpts, reference lists follow standardized styles, and procedural prose does not inadvertently mirror previously published protocols.

Ask CASRAI · free to tryQuestion about this guide? Ask CASRAI answers with linked sources. 2 free questions a day, no account.

Ask your first question freeSubscribe

Structure (ordered list)

A standard Unicheck similarity report presents the document alongside diagnostic panels and filtering tools that permit granular evaluation of text overlap. The report is organized into the following components:

  1. Similarity Percentage Indicator: The top-level summary displaying the overall proportion of matched text across the entire document. Scores are accompanied by color-coded visual indicators ranging from blue (0% matching text) and green (1% to 24%) to yellow (25% to 49%), orange (50% to 74%), and red (75% to 100%).
  2. Side-by-Side Highlighted Document Viewer: The central viewing pane displaying the manuscript with its native layout intact (retaining tables, figures, charts, and embedded images). Overlapping text passages are highlighted in source-specific colors that correspond to entries in the external source breakdown.
  3. Source Matching Panel: An expandable sidebar listing every external web URL, open-access repository, and institutional file matched against the manuscript. Each entry details the percentage contribution of that source to the overall score and provides direct links to inspect the source text.
  4. Automated Citation and Reference Exclusions: Algorithmic toggles that detect bibliography sections and in-text parenthetical citations formatted in American Psychological Association (APA), Modern Language Association (MLA), Harvard, Chicago, or Turabian styles. When enabled, references are marked in purple and in-text citations in blue, and the aggregate similarity index recalculates immediately upon exclusion.
  5. Matching Sensitivity and Filter Thresholds: Configurable filter parameters that omit small, ubiquitous phrases. The platform supports a 4-word minimum match sensitivity and allows reviewers to exclude sources contributing less than 5% total overlap, filtering out standard academic idioms and transitional phrasing.
  6. Character Replacement Diagnostic Panel: A security inspection layer that flags intentional manipulation attempts, such as replacing Latin letters with visually identical Cyrillic characters (homoglyphs) or inserting hidden whitespace to trick string-matching algorithms.
  7. Exportable PDF Summary Report: A downloadable summary document capturing highlighted passages, source attribution links, active exclusion settings, and final adjusted percentage scores for institutional archiving or editorial communication.

A short annotated example

To understand how exclusion filters alter a raw similarity percentage, consider an annotated example of a 2,500-word empirical manuscript submitted for pre-publication review:

  • Raw Initial Score (24% — Green/Yellow Threshold): The initial automated scan flags 600 words across four distinct categories before filters are applied:
    • Reference Section (275 words, 11%): Complete bibliographic entries formatted in APA style matching external repository records.
    • Direct Quotations (125 words, 5%): Two block quotations enclosed in quotation marks citing foundational theoretical definitions.
    • Laboratory Methodology (150 words, 6%): Standardized descriptions of buffer preparation, centrifugation speeds, and assay incubation times matching published laboratory protocols.
    • Introduction Section (50 words, 2%): A two-sentence string in the literature review mirroring an indexed article without quotation marks.
  • Application of Exclusion Filters:
    • Toggling “Exclude References”: The system identifies standard APA headings (“References”) and subtracts the 11% bibliographic overlap, bringing the score to 13%.
    • Toggling “Exclude Quotes”: The system recognizes quotation marks and removes the 5% quoted text, dropping the net similarity score to 8%.
  • Annotated Editorial Interpretation:
    • Methodology Overlap (6%): Accepted as conventional disciplinary phrasing; no rewording needed as long as primary protocol citations are present.
    • Introduction Overlap (2%): Identified as an unquoted near-verbatim passage. The author must rewrite these 50 words into original synthesis and verify citation placement to achieve an acceptable similarity percentage.

Common mistakes

Authors and reviewers frequently encounter interpretive pitfalls when evaluating Unicheck similarity outputs. Key errors include:

  • Treating Similarity Scores as Misconduct Verdicts: Assuming that any score above 0% indicates dishonest writing. High scores often reflect comprehensive bibliographies, extensive quoted passages, or statutory terminology rather than unethical appropriation.
  • Overlooking Citation Exclusion Settings: Evaluating unadjusted similarity percentages without verifying that automated APA, MLA, Chicago, Harvard, or Turabian reference exclusions are active.
  • Ignoring Small Match and Percentage Filters: Leaving the filter threshold at zero, which triggers false-positive highlights on standard 3-word or 4-word academic transitions (such as “in accordance with” or “the results demonstrated that”).
  • Inadvertent Self-Plagiarism on Draft Resubmission: Uploading preliminary thesis drafts to an institutional database with the “Sharing files” option active. Subsequent submissions match against the earlier draft, causing artificial 80% to 90% matches flagged as self-plagiarism.
  • Attempting Alphabet Homoglyph Substitutions: Replacing Latin letters with identical Cyrillic characters (such as “a”, “e”, or “o”) in an effort to bypass string matching. Unicheck’s character replacement algorithms detect alphabet hopping and flag the altered words with explicit warning banners.
  • Failing to Archive PDF Reports Before System Migration: Neglecting to download PDF reports of past assignments and thesis submissions before institutional LMS migrations (such as transitions from Unicheck to Turnitin Similarity following vendor deprecation).

Checklist

Use this pre-submission and review checklist to ensure proper manuscript evaluation in Unicheck:

  • Confirm document compatibility with supported file formats (DOCX, PDF, RTF, ODT, TXT, HTML, or Apple Pages).
  • Verify that tables, charts, mathematical expressions, and figures remain correctly positioned in the document viewer.
  • Enable automated reference and bibliography exclusions matching the active citation manual (APA, MLA, Harvard, Chicago, or Turabian).
  • Activate quotation exclusions to omit properly punctuated direct quotes from the similarity index.
  • Set the minor source exclusion threshold (e.g., omitting sources under 5% or matches under 4 words) to eliminate standard academic transitions.
  • Inspect all highlighted blocks in the literature review and discussion sections to ensure every match is paraphrased and cited.
  • Review the diagnostic panel to confirm zero character replacement or alphabet substitution alerts.
  • Ensure institutional repository deposit settings are configured appropriately for preliminary drafts.
  • Download and export the final annotated PDF similarity report for departmental or editorial records.

Discipline differences (only if sourced)

Similarity patterns and baseline thresholds vary substantially across academic fields due to methodological conventions and citation practices:

  • Physical Sciences, Chemistry, and Biomedicine: Experimental methodologies rely on standardized sequences, chemical formulas, and reagent designations. Under publisher standards, procedural overlap in methodology sections is routine and evaluated leniently, whereas overlap in the abstract, introduction, or discussion undergoes rigorous scrutiny.
  • Engineering and Computer Science: Publications adhering to technical standards—such as the IEEE Plagiarism Policy—maintain strict boundaries between reusable mathematical formulations or public algorithm pseudocode and original analytical narratives. Algorithmic code blocks require explicit documentation to avoid triggering false similarity matches.
  • Humanities and Qualitative Social Sciences: Scholarly papers feature dense textual analyses, extensive direct quotations, and historiographical commentary. Because qualitative arguments require original prose construction, unquoted similarity in introductory literature reviews or analysis sections is scrutinized more heavily than in empirical disciplines.

FAQ

What is Unicheck?
Unicheck is a cloud-based plagiarism detection and similarity checking platform originally launched by Phase One Karma as Unplag in 2014 and acquired by Turnitin in June 2020. It integrates with LMS platforms (such as Canvas and Moodle) to scan academic papers against indexed web pages and institutional repositories.

What happened to Unicheck after the Turnitin acquisition?
Following the June 5, 2020 acquisition by Turnitin, Unicheck continued operating as a standalone tool until reaching vendor end-of-life on January 1, 2025. Educational institutions migrated their plagiarism screening workflows from Unicheck to Turnitin Similarity.

How does Unicheck detect character replacement and alphabet substitutions?
Unicheck incorporates automated character replacement detection algorithms that identify letters substituted from different character sets—such as Cyrillic glyphs swapped into Latin words—highlighting modified characters and identifying the underlying source text.

Can Unicheck exclude bibliographies and quotations automatically?
Yes. The platform includes automated citation recognition for APA, MLA, Harvard, Chicago, and Turabian styles. When reference and quote exclusions are enabled, bibliographic lists and direct quotes are omitted, and the similarity score recalculates immediately.

Does a high Unicheck similarity score mean a paper is plagiarized?
No. A high similarity score indicates textual overlap with indexed material, which frequently reflects properly cited quotations, standardized methodology descriptions, or reference entries. Human review is required to evaluate whether the overlap represents legitimate citation or uncredited duplication.

Sources

  • Turnitin. “The Turnitin Family Expands.” Official Corporate Press Release. Published June 5, 2020. URL: https://www.turnitin.com/press/the-turnitin-family-expands
  • Moodle Marketplace. “Unicheck Plagiarism Checker Plugin (plagiarism_unicheck).” Plugin Directory and Documentation. URL: https://moodle.org/plugins/plagiarism_unicheck
  • Wayne State University Computing and Information Technology. “Plagiarism Review Migration to Turnitin Similarity.” Institutional Guidance. Published October 2024. URL: https://i.wayne.edu/view/6717c5f10011e
  • Wikipedia. “Unicheck.” Overview and Organizational History. URL: https://en.wikipedia.org/wiki/Unicheck

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Unicheck: Reading Similarity Reports

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →