Current location - Education and Training Encyclopedia - Graduation thesis - I found a paper on how to reduce the repetition rate on the Internet.
I found a paper on how to reduce the repetition rate on the Internet.
1, HowNet is detected segment by segment. When 13 identical words are detected, they are considered to be the same, so they are the same continuously, and should not exceed 13 words. Point out the similarities in the paragraphs and then combine them. On this basis, paperpass has a comparison function. Look at the number of words in this sentence that are similar or similar in length to other papers, and you can see the degree of similarity. For example, this sentence is 55% similar to another sentence. Wanfang began to detect the repetition rate in six words. When a paragraph is repeated a lot, every six words will turn red, and the detection mechanism is not bad. Unfortunately, the database is too small and has no advantage.

2, try to use synonyms instead, such as: damage = destruction; Channel = method; Principle = basic idea; Invisible = hidden; Outstanding advantages = obvious advantages, maximum = maximum, and so on. In the pre-detected report, the area marked in red is the key modification area. Take paperpass as an example, there are three colors, the green one doesn't need to be modified, and the yellow one doesn't need to be modified if it's not a large area, if it's just one or two sentences.

3. Changing the active voice and passive voice of sentences, such as digital watermarking, provides an effective technical means for multimedia data files in authentication, anti-counterfeiting, anti-tampering, and ensuring the security and integrity of data. Digital watermark detection can be used as a powerful detection method for authentication, anti-counterfeiting, anti-modification, transmission security and integrity assurance of multimedia products. 4. If there is a pause in the same sentence, or there are many aspects at the same level, you can adjust some orders. Here is an example. According to this degree of revision, it is guaranteed to pass:

Before modification: At present, digital watermarking technology has become the forefront of international information security academic research, which has opened up a new way for the safe preservation and transmission of multimedia. Digital watermarking ensures the ownership of the original data by embedding copyright information, that is, watermark, which can be a digital signal such as a paragraph of text, logo, serial number and image. Moreover, this kind of watermark is usually invisible, closely combined with the original data and hidden in it, and can be preserved through some operations that do not destroy the use value or commercial value of the original data. When necessary, the watermark can be extracted by certain technical detection means as evidence for judging the ownership of media copyright and tracking and prosecuting illegal infringement. Digital watermarking provides an effective technical means for authentication, anti-counterfeiting, anti-tampering of multimedia data files and ensuring the security and integrity of data.

Revision: At present, a research hotspot in the field of international information security is digital watermarking technology. This technology can provide a safer channel for multimedia storage and transmission. The principle of digital watermarking technology is to use watermarking to ensure the ownership of data, that is, to embed specific copyright information in the original data saved or transmitted. These embedded information, namely watermarks, can be in various forms, such as digital signals such as logos, characters, images and serial numbers. Moreover, these digital signals are invisible on the surface and closely combined with the original data in various forms. Of course, this combination will not destroy the original data, so the data will not be affected at all during use. If it is necessary to detect the watermark during use, the hidden watermark can be extracted by relevant means, and the copyright of the original data can be judged by this method, which can be used as evidence of legal copyright infringement. This technology has a broad application prospect in some fields, especially in the authentication, anti-counterfeiting, anti-modification and transmission security and integrity of multimedia data, products and documents.

In addition, the picture is undetectable. Some contents can be replaced by tables or pictures, or some tables or pictures can be converted into words, which can also reduce the repetition rate. There is also the use of references. No matter which detection method is used, the default citation method is in the form of endnotes, which is easy to identify. I put the method of endnotes in the following: 1. The method of inserting endnotes in files: 1, and put.

Enter "e" in the search field and replace it with: "[&; ]"。

2. If something like 12345 needs to be expressed as 1-5, you can select it and then use the keyboard shortcut: ctrl+shift+H to hide the selected content, manually enter 1-5 yourself, and then click CTRL+SHIFT+'+' to put it in the upper right corner ~~OK ~ ~

Revision suggestion

Our suggestion may not reduce your compliance rate to 0%, but at least it can greatly reduce your compliance rate and let you pass the written test of the school smoothly.

1. If the number of words in your paper far meets the requirements of your school for the number of words in your paper, then deleting some red words appropriately is the most direct way to reduce the coincidence rate.

2 .. The software will not recognize the words in the picture. So for some words that we can't modify at all and have to quote, we can turn them into pictures. Actually, there is a very simple method. Enter other people's things into your layout first, and determine the location and format. It is best to start from the end of the thread, that is, the whole block (quadrilateral). The size and font of this part of the text must be the size and font of the text you finally need to write. Then, copy your file, leave other people's things, delete all the rest and save it as another file on the desktop. This file will only contain other people's text content. Right-click to copy (remember the icon), don't go in, then open your paper and paste it where it is inserted. This method can ensure that other people's text content is a picture, but in the layout, you can't see that it is already a picture, exactly like your text, but you can't edit it. This method probably won't be seen in any books. Absolutely unique, simple copy and paste, but very knowledgeable. It is suggested that students use this method unless they have to quote it, but there is no way to modify it into other words (such as some concepts, proper nouns, etc.). ).

3. Don't choose documents for plagiarism. Some students may wonder why they clearly quoted or copied paragraphs or sentences from other documents in their papers, but they didn't detect them. This is because the sensitivity of the detection system of HowNet has set a threshold, which is 1%. In terms of paragraphs, plagiarism or quotation below 1% cannot be detected, which is common in clauses or small concepts in large paragraphs. For example, if the paragraph 1 has 10000 words, a single document that references 100 words or less will not be detected. In fact, here also tells the students a method of revision, that is, never choose an article to quote from paragraph plagiarism, try to choose as many documents as possible, and intercept a few words from one article, so that it will not be found out.

The most important way to modify the overlapping part is to translate the plagiarized content into your own language. For example, converting keywords into synonyms (HowNet detection system largely searches for keywords in sentences); Break long sentences into short sentences, and combine short sentences into long sentences; Change the way of description (change the original sentence into inverted sentence, passive sentence, active sentence, etc. ); Disrupt the sentence order in plagiarized paragraphs; Find more foreign materials or traditional Chinese characters, and those who have the ability can translate some foreign languages. Now there are many translation software, such as Baidu dictionary, and the translation results are very accurate. Traditional Chinese characters are relatively simple and can generally be understood. Just put it another way. But remember that simple addition and subtraction of words are not feasible in the detection system of HowNet.

5. For the quoted content, even if the label is correct, it will be included in the coincidence rate, so even when quoting, we'd better convert the quoted part into our own language to re-describe.

6. The detection of forms by this system is still not perfect. According to my observation, the overlapping words in the table are not included in the total coincidence rate. Therefore, what cannot be transformed can appear in the paper in the form of a table. 7. If you really don't know how to modify the quoted content when quoting a long document, the sentence order of the quoted paragraph will also partially reduce the coincidence rate, and it is better to modify it slightly in the process of confusion. However, it is impossible to eliminate all overlapping words through such modification.

Practice has proved that the above methods can reduce the replication rate by more than half and ensure the smooth passing of the exam. Of course, there is no guarantee that all the modified words will no longer be marked in red, but at least half of them will no longer be marked in red.