Plagiarism as a current problem

Tools that you discover, capture, detect plagiarism and also act as a preventive in the direction of other such cases it can be divided into several groups. In particular, the two large groups of instruments:

- also tools that are focused on working with the content pages of the document,

- the tools that the content pages of the document are not working, that do not perform document comparison in terms of content with each other.

Regarding the tools that work with the content of texts and works, it is possible to distinguish those that allow to compare a particular text or some parts of other documents (and their components) which form part of the comparative corpus. Depending on with what the text or its parts are compared, we distinguish:

- intracorporal text or parts thereof are compared with other documents in a single cabinet,

- extracorporal text or its parts are not compared with those documents which are part of a particular corpus, but compared with the documents contained in the corpus of others (ie compared with documents from other corpora),

- mixed these allow you to compare texts and documents not only within the same cabinet, but also with documents outside that form part of other corpora; thus combining intracorporal and extracorporal comparison.

Another group of documents that work with a work are called, internal tools. These work only at the level of a single document and are aimed at to determine whether the plagiarism using what is known analysis. Specifically, the document is thus subjected to this analysis, which is based on the assessment of the style in which the text is written. This style is actually the style of the author and any other styles thus suggests that parts of the text were used from other sources. Naturally examines whether the appropriated or been properly granted by reference. This is the method in which certain words removed from the text (eg. In order every fourth or fifth word) and then prompts the author to these words in the text added.

According to the thought and stylistic sequence of words is then determined by the number of those author of the report added, and the place and time of this down the percentage probability that output is plagiarized (resp. Degree of originality, taking over other knowledge and ideas).

The second group of tools are those that does not compare content page document (do not work with a job). Their essence is to work with the addition of parts, supplemented with data corresponding to the added texts and thus actually preclude the output of this activity was the nature of plagiarism:

- using invisible tagging (labeling) in electronic form that is passed on to the assessment and that is formed in accordance with the prescribed template are hidden and these particular brand within the insertion of other parts, other texts somewhat substantially degraded,

- use APE editor it's antiplagiarism editor, which has however relatively limited functions of copying and also for insertion into a particular document.

According to available materials, it is clear that at present in order to detect plagiarism effective functioning quite a few systems. Those currently operating effectively, have the character of tools that work with text and work carried out intraand also extracorporal comparison. They combine so comparison with the works contained in the considered corpus, but also text compared with other texts outside the body (eg. the Internet). If we should determine which system is the most efficient in terms of detecting attempts at plagiarism, it is necessary that the systems integrate the following elements:

1) must be able to actively seek,

2) must be able to search within your own website,

3) they must be suitable to compare the text and shall also be liable for the performance of other methods of artificial intelligence,

4) indexing data on the level of detail must be revised,

5) must take into account the structure and division of work into different areas, departments, according to research focus, by keywords, and the methodology used,

6) there must be a statistically must give also reduce texts and works that form the background material for the work being compared text.

Very often we meet also with the view that the essence antiplagiators systems is a psychological effect, creating a specific mechanism of supervision, the final work is inspected and evaluated. Student so basically gets the feeling that he was under some form of supervision and ultimately leads to more responsible approach and responsible working student. If this situation combined with some suitably publicly presented cases of plagiarism, which have been discovered, designed and ultimately affected and remedy in this area is relatively easy to create, respectively, at least in the right direction to focus the creative activity of students.

Very interesting are the statistics about how resources are used respectively. The plagiarism abuse. Students generally use the first of the sources found (the first three in the order), which after entering the title work, topics or keywords are searched the most commonly used browsers. According to this can be relatively quickly and efficiently check the transmitted texts in terms of their authenticity and originality.

For detecting plagiarism it is also important to understanding the psychological page forger. It is based on the fact that if someone is cheating in this way and are usually located in stressful situations and does not place such emphasis on the format, the same alignment, default languages, does not use the same words and even phrases.

2. Systems to detect plagiarism.

As I have already indicated that plagiarism is not only a problem encountered in the Slovak Republic. It is a globally widespread phenomenon, not only in education but also in areas that are associated with research and development, or in the arts (everywhere, where you can work and develop specific people abuse others so that no return shares issued to the original author's own, created by a person plagiarist).

It is possible to mention a few systems with which it is most often in the Internet environment, which are designed to detect plagiarism. Naturally, much of the systems and programs is already obsolete, and therefore further detail is not developed and not develop. On the other hand, we meet also with the systems and programs that work very well, reliably and include a relatively wide range of sources under consideration. From the busiest systems in order to detect plagiarism can mention the following:

1) the Tumitin.

2) the iTHENTICATE.

3) the Masaryk University UIS.

4) the CopyCatch.

5) the Urkunde.

6) the Ephorus.

7) the SafeAssign.

8) the PlagiarismDetect.

9) the Plagiarism-Detector.

10) eve2 system.

The first of the mentioned systems Tumitin's post leader in the provision of services that aim to identify and detect plagiarism. It is available on www.tumitin.com or www.plagiatorism.org while its owner is a company iParadigms. For this system it is typical that operates with its own data sources and also has its own source search. Is redesigned so that it can archive several tens ofbillions of websites, more than 70 million different work (drawn up students in education at various levels), archives also for book sources, magazines. In terms of language compatibility and usability is important, it allows the assessment in more than 30 languages including Slovak and Czech language. In the case of purchase, the buyer sent documents via the website (thus regards the provision of services).

The work of this system lies in the fact that for each document makes a certain impression, and this is then compared with a database of existing fingerprints. Naturally, the prints were kept classified and indexed. Imprint of a particular document, however, does not compare with the entire existing database. The criteria chosen performs system selection and compares sources (Their fingerprints) to the relevant group of database resources (the fingerprint). The system essentially continuously searching for relevant resources from the database (books, magazines, websites) And ongoing assessment document compares them.

This system is particularly widespread in the UK, where it works with most universities. Transposed is also in many Asian and American colleges and universities, while in our country it operates with one high school and college private nature College of Management (City University of Seattle).

The problem with this system at certain times of the operation was that it fundamentally violates copyright law students. These in fact commissioned work into it, but nepodpisovali any documents that would enable archiving works. Thus it was essentially a system filled with lots of work and still get the most out of this without being involved in the company's profit paid to students who commissioned the original work and that actually enable the development and effectiveness of the current system. In recent years, even it declared that the system allows the control of grammar, which was already the intention of the prior few years.

In addition to the problem of copyright and interference with intellectual property system also has problems with access to fee-based databases. In addition to that this system can be considered quite good and this both in terms of speed, stability, considered in terms of the amount of work (access database), in terms of sophistication and support.

While the system Tunitin is characteristic and typical of the academic field of education and a similar program it has also been developed for non-academic area. Namely the publishing sector, also can mention legal and governmental organizations and institutions. Within them it is used systemically and structurally similar to conformity assessment system and the system iTHENTICATE, owned by the same company.

The system used at Masaryk University, namely UIS system was developed in Brno and in terms of language compatibility it allows comparison of Slovak and Czech documents. It creates a custom database work and compares these with new, inserted texts. Likewise they compare them with Internet resources. Currently, the system is functioning as the National Registry of Theses and Plagiarism, which is funded by appropriations from its own Ministry of Education. Originally he served as an integral part of university information system at MU, but later on it was allocated. In terms of structure it consists of two components namely the National Registry, which serves the provision of registration of descriptive information about the final thesis and the other one is part of a system that allows the detection of plagiarism.

In terms of detecting plagiarism can be said that the continuity of individual steps is similar as with other systems. Certain particulars is the possibility of teacher participation in the process of comparison so that the process of comparing a known ability to add resources to the chosen topic work in the system. Thus essentially somehow it extends the range of sources which the work is compared. The actual comparison is carried out so that the text is broken down into several smaller units and these are followed by a certain algorithm compared with the available resources. Around mid-2008, the system was enhanced with new algorithms that should facilitate the processing of large amounts of data for a more detailed comparison. Likewise, existing algorithms improved to make them more efficient when searching for information, to be substantially more sensitive to the potential, the potential match.

Currently it works as www.theses.cz project, which involved more than 17 universities in the Czech Republic and considering the quality of it and show interest in our high schools. Among the Slovak you can mention for example. University of Economics in Bratislava, which is in the system of schools using the system involved first. plagiarism global research

Another relevant system is CopyCatch system that allows comparison of documents and individual parts. At present, its parent company Software Limited are several versions depending on the nature and requirements of the end user (in their different groups). From the perspective of users it is one of the best comparative systems that enable the detection of plagiarism. In terms of its technical evaluation but a lot worse. The main shortcomings were identified particularly problematic reliability and also that it likely can not cope with massive pressure that would, for example, in relation to the comparison work at universities clearly assumed. Also outputs (response) of the system are not provided by operational, and thus a further deficiency is considered to be a problem with the speed of response to your request. And ultimately it has already considered the lack and that this system works (it was created and works) in Java and in terms of support is limited exclusively to Windows.

Scandinavian Urkund system is currently the most widespread in those Nordic countries and in the nature of Web services and as such is integrated into other systems. In terms of efficiency and functionality when comparing documents and detecting plagiarism has only average results. Disclosure of the comparison is the same as with other systems, thus assessing the text or specific parts of the documents contained in a specific corpus. In view of this, identify the sources of work and it is often compared, it is important to note that the system takes into account both online resources as well as book publications, has published magazines and also the work of other students that are entered in the system.

Another of the assessment of only an average system is the Ephorus (available on www.ephorus.sk/home). In terms of comparison it operates in principle as above mentioned systems. It also has the nature of Web services, which controls both material compared to works that are entered in the system, but also to the Internet source. In terms of use it is most prevalent in the United Kingdom, the United States, Belgium, the Netherlands, and is also used by the University of Economics in Prague.

The SafeAssign (available on www.safeassign.com) developed by Blackboard. It is characterized by working with your own data sources, the comparison can search both the Internet system ProQuest as well as global reference database. It works on the same principle as the above systems as well as its output is very similar to that of the aforementioned systems it's actually a detailed view of a match. Compared with other systems it is an interesting example, that as students as well as teachers can ask to be checked some documents removed from the database.

PlagiarismDetect system is freely accessible (ie online) solution to the problem of plagiarism. Is it interesting example in that it provides essentially only interface and service that leads to control document. None of the documents, however, does not.

On the opposite principle than most programs operating system Plagiarism-Detector. In order to be able to work with this system, it is necessary to install a specific client, which is a processor of a document (work). The document is processed to search parts, and these are then evaluated in terms of the incidence of repetition, in terms of identification, assessment, for a total for the occurrence of plagiarism. The system uses Google's database and sends it to the individual blocks of text for the assessment. This system can be regarded as a classical system used in order to detect potential plagiarism. The fundamental problem, however, is that it has resolved contractual and general legal relationship with Google-om and its database, which can be in the future and existential problem.

For all forms of Windows is designed antiplagiarism system eve2. It works with Internet search engines and for the assessment of effectiveness and efficiency in detecting plagiarism is only an average system. The issue is the functionality in the mass exploitation and comparing that with final thesis in education basically assumed and also problems with speed communications and also in issues of reliability of the system are quite large reserves.

Finally advantages and disadvantages of available comparative systems. Already above outlined the system for the comparison of conformity of documents have certain advantages and disadvantages as well. It is possible to speak about certain restrictions on the applicability of certain limits their effectiveness in dealing with plagiarism. Perhaps the biggest problem is that different systems are used for comparison purposes various tools, techniques, methods, analyzes of other documents, as well as different algorithms, results of different systems in one and the same document in terms of level of originality, respectively, depending on the extent to which the document is plagiarized, may vary. In addition to this, it is necessary to take into account other problems encountered in using them:

1) all systems can compare the documents or parts of certain materials available under utilized databases contained by the source; as such, but still not reveal the extent or degree of plagiarism, but merely suggest abusiveness certain parts of the document or documents as such and the suspiciousness is a need to further examine and thus confirm or deny,

2) it is questionable whether and to what extent it is possible autoplagiarism considered as a form of plagiarism; Indeed, it is necessary to determine to what extent it is possible to use their own resources to process additional works and texts (how much, what percentage of those original can be used within other works are to be considered as original, original work); The system evaluates the autoplagiarism and therefore it is necessary to determine to what extent it is permissible (percentages),

3) the documents to be entered for comparison, are compared to a particular database of documents, books, articles, Internet resources, and the like; in the final stages of labor and professional resources, so it is natural that with Increases in the number of materials with which to compare the document, automatically increasing the percentage likelihood that detects plagiarism; the other part of the saturation decreases and the speed of response to the submitted request; also systems that do not work with their own repositories, you must essentially over and over again in any case to search all available resources and thereby the more the longer it takes naturally also obtain the output from this comparison,

4) documents and parts are compared only with those texts that are entered in the database. They cannot therefore determine compliance with the resources that they do not know (who in the database are not):

- with resources that are only on paper and not in electronic form it is essentially impossible to determine whether the entered work with them (documents or parts of them) agree,

- have systems that perform comparisons only within his own cabinet, I can identify potential plagiarism therefore unable to assess compliance with the documents and other sources that are outside the body, which is compared with the work,

- also on the contrary, have systems that compare papers (documents or parts thereof) of resources out comparative corpus, I cannot assess compliance with the resources contained within the body,

- a common problem is the unavailability of paid databases, which are not assessed compliance (many systems do not have access paid corpora).

5) documents that are intracorp instruments cannot be assessed and distinguished in terms of originality and originality of authorship; when compared, so both must be considered as a potential, possible plagiarism,

6) even if it is in two or more documents in the same corp it might seem that could be similar to each other or directly identical (if the thus appeared to be forgeries), this may not ultimately be; the solution of this case because it may be the source of two or more documents that match, may be different and may be located outside the body of the content,

7) some internal tools can reveal plagiarism only if the document contains sections of text that come from different authors; Tools in essence, that the work is more styles of writing which are written in different parts; if there even suspected, so I cannot answer the question, whence, from what source the parts of the text took, a potential source or do not know may not be able to find,

8) the problem remains that the existing systems are unable to assess compliance to the texts and the works that are in a foreign language (if the comparison with them have been translated into another language, although it could also be an absolutely literal translation),

9) systems cannot detect paraphrases, thus partially altered lyrics,

10) many systems cannot operate so as to clearly distinguish the identical parts of texts that have been duly cited, those cited, no wrong and those that were not cited at all; all of them considered as compliance and thus need to be followed closely examine the output from the system and re-checked and evaluated for possible pla

11) giarism; thus essentially somehow it distorts score plagiarism, although in real terms plagiarism does not go,

12) the unavailability of the network is often fatal problem on some systems,

13) in virtually every system to detect plagiarism has certain weaknesses and their possibilities, such comparisons can be bypassed; instructions such proceedings in some cases located directly on the Internet.

References

1. Classifications of Plagiarism Detection Engines. Lancaster, T. a Culwin, F. 2/2005.

2. Farrington, D., Palfreyman, D. The Law of Higher Education. Oxford: Oxford University Press, 2006.

3. Stephen J., Rosenwasser, D. Plagiarism.

Размещено на Allbest.ru