Android-based English Grammar Checker with Voice Recognition

Android-based English Grammar Checker with Voice Recognition ABSTRACT In this paper, the Researchers describe an Android-based Voice Recognition system that acts as a grammar checker specifically geared to the needs of non-native speakers who are willing to learn and be familiarized with English language. This paper also examines how wrong usage of grammar can affect the communication of people both orally and in writing. Most commercial grammar checkers on the market today are meant to be used by native speakers of a language who have good intuitions about their own language competence. These tools were made to help native speakers who were not fluent in other language, thus making their outputs grammatically correct. However, these errors vary from one another, the reason why different grammar checkers are designed for different specific needs. The grammar checking component uses island processing rather than a full parse. This approach is both rapid and appropriate when a text contains many errors. In this research, it aims to develop a complete voi ce-based English grammar checker application that will help non-native speakers that want to learn and to speak English fluently. The focus of this paper is the grammar checking component. This approach is both useful and more efficient, since it directly corrects the grammar by translating the voice into text. The system also used Internet Search Engines that will provide examples of how the content of the voice into text segment can be expressed in a grammatically correct and idiomatic way. This application can help people having a hard time in expressing themselves using the English Language. KEYWORDS Voice Recognition, Grammar Checker, Android, Natural Language Processing, Mobile, Application, Vocabulary, Grammar INTRODUCTION One of the common reasons of miscommunication is due to wrong choice of words, resulting to poor grammar. Poor grammar later results of miscommunication or worse, someone with bad grammar was often insulted or shamed by people. To avoid these, word-processing systems are developed at the present time already include grammar checkers that are used to locate different grammatical errors in a text. These tools were made to help native speakers who were not fluent in another language, thus making their outputs grammatically correct. It is also intended to help its researchers to write texts in English, which is not their language. Although their command of the language is generally acceptable, most of them do not feel confident about their correctness and the expressiveness of their writing. However, these errors vary from one another, the reason why different grammar checkers are designed for different specific needs. Internet provides a boundless number of documents in English so the main function of the checker is the use of an Internet search engine that detects the text segments that are not found on any web pages. But most of these tools were aimed to help native speakers to check their English grammar, but what about Filipino and other citizens who are not fluent in speaking in English Language? How can one point out his or her grammar mistakes when they are trying to write a paper in our language? We designed an Android application for grammar-checking in English Language by using voice recognition. This android application aims to help the users that wants to learn English Language fluently and at the same time, it can help to improve their English vocabulary and grammar. The system will require the user to record a voice message. After the recording, the system will translate the message into text. There will be two buttons, the record again and go. If the text output is wrong, the user should press the record again button and if the text output is correct, the user can proceed. The system will now check the grammar. For example, the user recorded â€Å"Matt like Fish†. The text output was correct, so the user pressed go. The system will now process the grammar checking and after a few seconds, the system will give the corrected result via text output which is â€Å"Matt likes Fish.† In order to find out the common errors usually done by users, we conducted a survey among potential users around the New Era University, specifically some Filipino students who are a bit struggling when it comes to speaking English fluently. We ranked the results by percentage so that we know what we should be focusing on. A total of 30 respondents participated in the survey. DESCRIPTION OF THE COMPONENTS Internet search engine Grammar Checker Internet Search engine The checker used the Wordnet 3.0 engine, a lexical database for the English Language. It arranges English words into sets of synonyms called synsets. The engines used to find the search results for a text segment are Google and Yahoo. Grammar Checker Part of a system that verify written text for grammatical correctness. The implementation of a grammar checker makes use of Natural Language Processing. Once the grammar checker has been active, the text which is based on the voice recorded by the user will undergo a number of stages. It is separated into sentences and words. The individual words are then looked up in the internet. It includes all the words that occur in the system with all its possible usage. METHODOLOGY The table shows the results of the survey. With this table, we are now informed about the common mistakes that the potential users might commit. This table will help us to distinguish and to focus more on the most common error to help the users more efficiently. The common error that had the highest percentage is the improper structure of English sentences. Next is the improper use of verbs. Third is an improper classification of Noun. Fourth is the improper combinations of certain words, fifth is incorrect use of Adjectives and lastly, is improper use of Adverbs. This application will help the users construct English sentences which are grammatically correct. In this way, the users will learn their mistakes and at the same time, learn how to speak and write in English fluently. The following table gives the percentages of errors found in the surveys conducted. Table 1. Error Percentage SYSTEM ARCHITECTURE Figure 1 Grammar Checker Architecture Figure 1 shows the architecture of the system. It shows the flow of the system and how the system works and also how to gain and give a possible output of the application. Figure 1.2 Voice Recognition In this picture, it shows the stress of the user’s recorded voice and translates it into text. The system will recognize the voice and checks its grammar in the process. After that, the user will now find out his/her mistakes and the system will provide the correct grammar. METHODOLOGY RELATED WORKS Automatic voice recognition has a long history speech processing. Automatic speech recognitioncan bedefinedas the independent, computer-driven transcription of spoken language into readable text in real time (Stuckless, 1994). There is a lot of research has been done for Android applications including Speech Analysis and Natural Language Processing. Voice Recognition today is still improving as additional uses for the technology are developed. The applied voice analysis to conclude activities happened around a user. Voice Recognition Technology can set up to depend on tools known as grammar models that help minimize the data required to recognize the signals. DESIGN AND ALGORITHM ANALYSIS The eCheck: An Android-based English Grammar Checker Using Voice Recognition is designed to guide and help not only Filipino citizens, but also other people that want to learn English fluently and at the same time, improve their English vocabulary and grammar skills. SYSTEM OVERVIEW As stated in this study, the Researchers applied the same basic principles as they have adopted before, however, the Researchers considered a chart to show the flow of the system and to show how our application work. The Researchers also tried to measure the connection of each sentence in the message of the set of categories in our study. Once the grammar checker has been activated, the text which is based on the voice recorded by the user will undergo a number of stages. It is separated into sentences and words. The individual words are then looked up in the dictionary (see Burnage 1990). It includes all the words that occur in the system with all its possible usage. There are some words that are the same, but has a different meaning. In this sense our dictionary is not a simplified type. It is vast but simply a shorter version. CONCLUSION / RECOMMENDATIONS The capability and reliability of the Voice Recognition technology have been ever-growing. With more attention and implementation into the world’s most common operating devices, this technology will continue to develop. The Overall evaluation of this application is very helpful, especially to the Filipino citizens. This application can generate data that can give answers to their grammar and speaking concerns, With the advancement of technology, doing a grammar check and spelling is easier than you think. It no longer requires hitting the books and having lots of thesauruses or dictionaries at hand. The Researchers recommend this application for non-native speakers having a hard time in expressing themselves using the English Language.This application can be used for the business and transactions that requires them to speak in English. Because this application can help them to learn more about the English Language by just speaking and by knowing the grammar proper and the right pronunciation of it. There are many features of English that make it fun and interesting to learn. The Researchers also recommend this application to Everyone who wants to learn the proper grammar in English, for them to improve, to gain more knowledge and to fully develop their speaking and also grammar concerns in their own language. By using this application, they are just like having a tutor and a grammar book. The Researchers also recommend that if this study continues to improve, it is needed to be developed and implemented because these tools were made to help native speakers who were not fluent in English language and to discover more about it.