This article summarizes the research on computational methods to detect academic plagiarism by systematically reviewing 239 research papers published between 2013 and 2018. To structure the presentation of the research contributions, we propose novel technically oriented typologies for plagiarism prevention and detection efforts, the forms of academic plagiarism, and computational plagiarism detection methods. We show that academic plagiarism detection is a highly active research field. Over the period we review, the field has seen major advances regarding the automated detection of strongly obfuscated and thus hard-to-identify forms of academic plagiarism. These improvements mainly originate from better semantic text analysis methods, the investigation of non-textual content features, and the application of machine learning. We identify a research gap in the lack of methodologically thorough performance evaluations of plagiarism detection systems. Concluding from our analysis, we see the integration of heterogeneous analysis methods for textual and non-textual content features using machine learning as the most promising area for future research contributions to improve the detection of academic plagiarism further.

https://dl.acm.org/doi/pdf/10.1145/3345317

Academic Plagiarism Detection: A Systematic Literature Review

The efficiency and scalability of plagiarism detection systems have become a major challenge due to the vast amount of available textual data in several languages over the Internet. Plagiarism occurs in different levels of obfuscation, ranging from the exact copy of original materials to text summarization. Consequently, designed algorithms to detect plagiarism should be robust to the diverse languages and different types of obfuscation in plagiarism cases. In this paper, we employ text embedding vectors to compare similarity among documents to detect plagiarism. Word vectors are combined by a simple aggregation function to represent a text document. This representation comprises semantic and syntactic information of the text and leads to efficient text alignment among suspicious and original documents. By comparing representations of sentences in source and suspicious documents, pair sentences with the highest similarity are considered as the candidates or seeds of plagiarism cases. To filter and merge these seeds, a set of parameters, including Jaccard similarity and merging threshold, are tuned by two different approaches: offline tuning and online tuning. The offline method, which is used as the benchmark, regulates a unique set of parameters for all types of plagiarism by several trials on the training corpus. Experiments show improvements in performance by considering obfuscation type during threshold tuning. In this regard, our proposed online approach uses two statistical methods to filter outlier candidates automatically by their scale of obfuscation. By employing the online tuning approach, no distinct training dataset is required to train the system. We applied our proposed method on available datasets in English, Persian and Arabic languages on the text alignment task to evaluate the robustness of the proposed methods from the language perspective as well. As our experimental results confirm, our efficient approach can achieve considerable performance on the different datasets in various languages. Our online threshold tuning approach without any training datasets works as well as, or even in some cases better than, the training-base method.

/pdf/scalable-and-language-independent-embedding-based-approach-43sw8n81h8.pdf

Scalable and language-independent embedding-based approach for plagiarism detection considering obfuscation type: no training phase

Plagiarism is generally defined as literary theft and academic dishonesty. This considered as the serious issue in an academic documents and texts. There are numerous of plagiarism detection techniques have been developed for various natural languages, mainly English. In this paper we investigate and review the plagiarism detection techniques and algorithms which have been developed for Arabic Script Languages (ASL), and providing a literature review of the utilized methods in terms of techniques and outcomes.  The result of this paper will help the researchers who are going to commence their development and extend their researches in ASL like Arabic, Persian, Urdu, and Kurdish.

/pdf/plagiarism-detection-techniques-for-arabic-script-languages-2r66ohgzsu.pdf

Plagiarism Detection Techniques for Arabic Script Languages: A Literature Review

Various approaches have been implemented for plagiarism detection used, for author‘s work and academic publication, there is a purpose to create such reliable and performant plagiarism detection with increasing amount of publications. This is a serious offense where one author presents someone else’s work as his own. Moreover these algorithms don’t consider similar sections for efficient comparison. The proposed framework performs efficient sections wise plagiarism detection and provides suggestions for improving documents. The precision, recall and accuracy based on different n-gram features is presented showing the strictness of higher level n-gram features.

/pdf/an-enhanced-framework-for-extrinsic-plagiarism-avoidance-for-1wot31945l.pdf

An Enhanced Framework for Extrinsic Plagiarism Avoidance for Research Article

In recent years, the rapid increase of Persian electronic resources and facility of access to them has seriously triggered the plagiarism problem of the Iranian scientific community. Despite the automatic systems of plagiarism detection, like Turnitin, Eve2, this problem has strongly remained due to lack of support from Persian. The main purpose of this article is to detect exact plagiarisms and re-writings in Persian science texts. In our proposed method, after the candidate retrieval based on the statistical characteristics, in the text alignment step, structural analysis and semantic analysis of expression has been performed to detect re-writing plagiarisms. Firstly, data-driven dependency parser has been improved with the help of a deep learning model for Persian language to analyze the structure of the expression, and then the degree of structural similarity of the expression is evaluated through the analysis of the dependency tree. In this paper, our suggestion to examine the semantic similarity of expression is to use the semantic role labeling obtained from the deep learning model presented. The experiments have been performed on the corpus prepared in the AAIC2015 and corpus of the PAN2015 competitions. The results indicate that structural and semantic information improves the performance of the proposed method. ParsiPayesh is available on http://www.parsipayesh.ir.

ParsiPayesh: Persian Plagiarism Detection based on Semantic and Structural Analysis

This paper presents a new approach for Persian plagiarism detection. This approach uses a graph structure as well as one of the graph similarity methods (iterative methods) for similarity detection of two Persian documents. In this approach, documents are represented by a graph with specified length, then each part of suspicious document is compared to that of the source document. The graph is made if these parts have more common bigrams than a predefined threshold. Once graphs are made, an iterative method is used to find analogous nodes in graphs. Two graphs are marked as similar if they contain at least a certain number of similar nodes. In order to evaluate the proposed method, it was run on PAN2015 and PAN2016 Persian Text Alignment dataset. The Plagdet score is defined to evaluate plagiarism detection methods in PAN contest. The gained Plagdet of proposed method is 90% on PAN2015 and 87% on PAN2016. CCS Concepts • Information systems➝ Plagiarism Detection software • Computing methodologies➝ Graph-based

/pdf/graph-based-approach-to-text-alignment-for-plagiarism-52vyeyo5zs.pdf

Mozhgan Momtaz

Papers

Graph-based Approach to Text Alignment for Plagiarism Detection in Persian Documents.