The Construction Of The Prototype Of The Automatic Correction System For Middle School English Composition

Benefit from the development of Internet technology,every field is taking great transformation nowadays.The modem medical devices,mobile communications,transportation,and smart homes are all showing the era of the Internet has been coming.Many information products also come forth in the field of education,like learning machines,smart tablets,early education robots,school systems and so on.All of them are promoting the development of educational information.From a investigation on the educational products,we found that the market exists a large mount of software for English-study,but rarely no one especially designed for English composition improvement.Therefor,the paper introduces the design and development of a system prototype for automatic correction of composition,aiming to satisfy the emergent demand of the market and hopes to help improve writing levels of English learner.The system is architectural with B/S model and developed by PHP program language and MySQL database,serves personal information management,composition correction,trouble shooting,history records viewing and other functions,in which the composition correction is the most difficult part in the architecture.In order to implement the function of composition correction,the data resources such as the outline vocabulary,example sentences and corpus are collected,and the Python language is used to implement the text processing algorithm.The text similarity,n-gram language model,and recommendation system are mainly applied.Text similarity and n-gram models are relatively mature technologies in the field of natural language processing;text similarity can use the edit distance and cosine distance to calculate the similarity between a sentence and the correct example sentence,and then judge the difference from the correct sentence.The n-gram model compares sentences based on a statistic of the probability of adjacent words in a large amount of text,and the system to improve the probability model,not only calculate a certain The probability of a word appearing after a word,and the probability of a word appearing in front of a word is also calculated.Using the combination of the two results in a better model;and the recommendation system is the system that determines the error in the composition sentence.Finding similar correct sentences is recommended for learners' reference.Each treatment method has its own advantages and disadvantages.When the system is developed,a variety of processing technologies are used together to achieve better results and achieve the desired goals.