SzTextCompar

Open two text/document files side by side, with synchronized scrolling and a line-number gutter.
Nine independent analysis methods, each giving its own similarity score and a plain-language verdict:
Traditional Diff (character/line level)
Winnowing Fingerprint — robust to reordering and small edits
TF-IDF Weighted Content Analysis (precise cosine similarity, rare words weighted higher)
Sentence Rhythm & Structure — a paraphrase filter based on sentence-length patterns
N-gram Phrase Matching (common 4-word sequences)
Vocabulary (Cosine) Similarity
Jaccard Similarity (set-based)
Simhash Fingerprint (fast, language-independent)
Summary Table & Similarity Index — a weighted overall score combining all of the above, with a table and a chart
Visual Match View: highlights every matching phrase (configurable minimum word count) in both texts with a color-coded palette, with a clickable match list and next/previous navigation.