improved text mining approach for plagiarism detection

Abstract

newline Plagiarism causes significant damages on innovations in scholarly works such as student newlineassignments and research. It appears in various forms like Accidental, Unintentional, newlineIntentional and Self plagiarism and spans across varied data formats: text (articles, scientific newlinepapers , essays. or just plot ideas in fiction books,Video (scenes, dialogues), music foms newline(melodies,tunes,compositions ), images (statistics, charts graphics and drawings),internet newlinesource streams. Text plagiarism comes in multifaceted forms: low-obfuscation (verbatim, newlineparaphrasing and its variants) as well as high-obfuscation (semantic and conceptual or idea newlineplagiarism) and more. newlineThe state-of-the-art and commercially available plagiarism detection merely detect pairs of newlinesections bearing similar text chunks, both from suspicious and source documents, and hence, newlinehandling low-obfuscation plagiarism tasks. This leaves us at exploring high obfuscation newlinelevels of plagiarism as one of the research challenges. The current research objective aims at newlinedeveloping a novel method of plagiarism detection tool based on complete and precise newlineguidelines heuristics for anti plagiarism standards both mapped from national and newlineinternational level. The tool was designed in form of a text semantic frame work to reveal the newlinetype of plagiarism in the form of word-to-word, paraphrasing, syntactic, semantic or idea newlineplagiarism. It incorporated logistics for improving the accuracy of existing plagiarism newlinedetection approaches using semantic based Natural Language Processing (NLP) techniques newlineupon free text available at different levels of granularity viz. documents, paragraphs and newlinesentences. newlineThe research began with drafting the guiding heuristics on all plagiarism detection categories newlineand mapping them to a precise set of IF-THEN Rules in order to type-cast to an exact newlineplagiarism category. The second part of the research focused on detection of idea plagiarism, newlinestill a challenging task for plagiarism detection research community. The outcome of suc

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced