Student Submission URLs
Enter each student's published internship website URL. Fields auto-expand up to 10.
i
Similarity Threshold
The minimum score a pair of students must reach to be flagged as suspicious.
Computed as: Jaccard Γ 55% + Cosine Γ 45%, plus structural bonuses.
40β55%
High recall β many flags
65% β¦ default
Balanced β recommended
75β85%
High precision β strict
i
N-Gram Size
The length of shared word sequences used to detect copied text.
Smaller sequences catch more matches; larger ones require longer verbatim passages.
3-grams
Sensitive β catches paraphrasing
5-grams β¦ default
Balanced β recommended
7-grams
Strict β verbatim only
Cross-Student Similarity Matrix
Pairwise text similarity scores. Red = high similarity = evidence of copying.
Individual Audit Reports
Per-student analysis with evidence artifacts and risk scoring.
Matched Passages
Exact or near-exact text blocks found across multiple submissions.
Generate Formal Evidence Report
Produces a print-ready, timestamped HTML document suitable for disciplinary proceedings.