Plagiarism can be of many different natures, ranging from copying texts to adopting ideas, without giving credit to its originator. This paper presents a new taxonomy of plagiarism that highlights differences between literal plagiarism and intelligent plagiarism, from the plagiarist's behavioral point of view. The taxonomy supports deep understanding of different linguistic patterns in committing plagiarism, for example, changing texts into semantically equivalent but with different words and organization, shortening texts with concept generalization and specification, and adopting ideas and important contributions of others. Different textual features that characterize different plagiarism types are discussed. Systematic frameworks and methods of monolingual, extrinsic, intrinsic, and cross-lingual plagiarism detection are surveyed and correlated with plagiarism types, which are listed in the taxonomy. We conduct extensive study of state-of-the-art techniques for plagiarism detection, including character n-gram-based (CNG), vector-based (VEC), syntax-based (SYN), semantic-based (SEM), fuzzy-based (FUZZY), structural-based (STRUC), stylometric-based (STYLE), and cross-lingual techniques (CROSS). Our study corroborates that existing systems for plagiarism detection focus on copying text but fail to detect intelligent plagiarism when ideas are presented in different words.

/pdf/understanding-plagiarism-linguistic-patterns-textual-1ysubwdzsv.pdf

Understanding Plagiarism Linguistic Patterns, Textual Features, and Detection Methods

A question answering system includes a question answering unit, an analysis unit, a tree structure generation unit, a feature extraction unit, an evaluation unit and a re-ranking unit. The question answering unit executes search processing based on an input question and acquires answer candidates to the question. The analysis unit executes syntactic analysis processing or syntactic and semantic analysis processing on a passage obtained as a result of the search processing. The tree structure generation unit generates a tree structure based on an analysis result. The feature extraction unit extracts a relation between a search word applied in the search processing and each acquired answer candidate as a feature corresponding to each answer candidate from the tree structure. The evaluation unit determines an evaluation value of each answer candidate based on the feature extracted. The re-ranking unit re-ranks the answer candidates based on the evaluation values.

Question answering system, data search method, and computer program

Word similarity is broadly used in many applications, such as information retrieval, information extraction, text classification, word sense disambiguation, example -based machine translation, etc. There are two different methods used to compute similarity: one is based on ontology or a semantic taxonomy; the other is based on collocations of words in a corpus. As a lexical knowledgebase with rich semantic information, How-net has been employed in various researches. Unlike other thesauri, such as WordNet and Tongyici Cilin, in which word similarity is defined based on the distance between words in a semantic taxonomy tree, How-net defines a word in a complicated multi-dimensional knowledge description language. As a result, a series of problems arise in the process of word similarity computation using How-net. The difficulties are outlined below: 1. The description of each word consists of a group of sememes. For example, the Chinese word “暗箱(camera obscura)” is described as: “part|部件, #TakePicture|拍攝, %tool|用具 , body|身”, and the Chinese word “寫信 (write a letter)” is described as: “write|寫, ContentProduct=letter|信件”; 2. The meaning of a word is not a simple combination of these sememes. Sememes are organized using a specific knowledge description language. To meet these challenges, our work includes: 1. A study on the How-net knowledge description language. We rewrite the How-net definition of a word in a more structural format, using the abstract data structure of set and feature structure. 2. A study on the algorithm used to compute word similarity based on How-net. The similarity between sememes, that between sets , and that between feature structures are given. To compute the similarity between two sememes, we

/pdf/ji-yu-zhi-wang-de-ci-hui-yu-yi-xiang-si-du-ji-suan-word-3q2xf1mp2c.pdf

基於《知網》的辭彙語義相似度計算 (Word Similarity Computing Based on How-net).

In this paper we describe a cluster-based plagiarism detection method, which we have used in the learning management system of SCUT to detect plagiarism in the network engineering related courses And we also used this method to detect external plagiarism in the PAN-10 competition The method is divided into three steps: the first step, called pre-selecting, is to narrow the scope of detection using the successive same fingerprint; the second step, called locating, is to find and merge all fragments between two documents using cluster method; the third step, called post-processing, is to deal with some merging errors Our method ran 19 hours in the PAN-10 competition, and the result ranked the second place, which met our expectation Keywords Plagiarism detection, Similar text, Locating, Cluster

/pdf/a-cluster-based-plagiarism-detection-method-lab-report-for-2n7jvvq9qd.pdf

A Cluster-Based Plagiarism Detection Method - Lab Report for PAN at CLEF 2010.

Plagiarism has become one area of interest for researchers due to its importance, and its fast growing rates. In this paper we are going to survey and list the advantage sand disadvantages of the latest and the important effective methods used or developed in automatic plagiarism detection, according to their result. Mainly methods used in natural language text detection, index structure, and external plagiarism detection and clustering -- based detection.

Survey of Plagiarism Detection Methods

In order to reduce the dimensionality of feature vector space and reduce the computing complexity of categorization, each document of the train set is summarized automatically and two approaches to text categorization based on these summaries are proposed: in the first approach, the text summarization is directly used for feature selection and categorization instead of the original text; in the second approach, each summary is used to select and weight features for each document, and free texts are classified using KNN algorithm. Experimental results show that the two proposed methods using automatic summarization can not only reduce the time of classifier training, but also improve the performance of text categorization.

Improving the Performance of Text Categorization Using Automatic Summarization

Aiming at the Chinese academic paper plagiarism detection, proposed chunk based plagiarism detection algorithm with chunk extraction method based on character or word. Taken account of that different part of document has different importance, proposed two paragraph weight algorithms and defined three paragraph weight functions. The best chunk lengths are determined by experiments. Experiments show that using paragraph weight can enhance the detection effect.

Plagiarism detection in Chinese based on chunk and paragraph weight

The sentence similarity computation plays an important role in each field of Chinese information processing. A new approach to compute the Chinese question semantic similarity is presented, which is divided into two steps: the first step is to extract the question semantic representation from the question, and the second step is to compute the question semantic similarity based on the question semantic representation. This paper uses the method of question semantic model matching to extract the question semantic representation from the question. The experimental results show that the proposed algorithm worked more reasonable in the real calculation. In addition, our approach is of great value for many potential applications in the future.

A New Approach to Compute the Semantic Similarity of Chinese Question Sentence

This work brought forward a kind of arithmetic of information retrieval, namely combining the positive genes of genetic algorithm and vector space model on the base of nature language. Genetic algorithm is used for a predication case-frame of query in this system. Based on HowNet, this algorithm gains inherent characteristics of data objects, and retrieve the useful information according to those characteristics. So the system implements information retrieval from hierarchical knowledge of concepts. This work also introduces the key technology of information retrieval. According to the research on the algorithm model, we design an IR system in financial domain. It is the most effective particularly in question-answering system. In addition, it can be extended to the other domain. It considerably enhances the intelligence degree of information retrieval by this algorithm.

http://ieeexplore.ieee.org/iel5/9459/30108/01382377.pdf

Research on natural language IR system based on genetic algorithm and VSM

Automatic extraction of domain terms is the basis of domain ontology learning. General linguistic resources such as WordNet and HowNet can be applied to extract only partial domain terms from domain unstructured texts. In this paper, we firstly extract partial terms by calculating domain relatedness between words by HowNet. Then the extracted terms are semantically clustered with fuzzy c-means clustering algorithm based on properties of rough sets. Finally more domain terms are extracted from unknown words according to the clustering results with the method of machine learning. The experimental results showed that the method can not only extract domain terms as more as possible, but also ensure higher precision.

Xiao-Zhong Fan

Papers

Improving the Performance of Text Categorization Using Automatic Summarization

Plagiarism detection in Chinese based on chunk and paragraph weight

A New Approach to Compute the Semantic Similarity of Chinese Question Sentence

Research on natural language IR system based on genetic algorithm and VSM

Research on Method of Extracting Chinese Domain Terms Based on Rough and Fuzzy Clustering