[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86249-en":3,"doc-seo-86249-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},86249,687197207919,"Theodora","https://ap-avatar.wpscdn.com/avatar/a000253d6f5f7c60be?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779446848396160552",6,"Technology","Dzongkha Next Word Prediction System","Dzongkha is Bhutan’s national language, and official documents and literature rely on its complex script to preserve cultural value. However, Dzongkha typing is slow and error-prone because each syllable requires multiple keystrokes and efficient typing tools are limited. A Dzongkha Next Word Prediction System is proposed to reduce keystrokes by predicting likely next syllables/words. Using a dataset of 100,000 sentences and 28,344 unique words, the pipeline preprocesses text, builds N-gram sequences, and trains LSTM, Bi-LSTM, and GRU. GRU achieves 74.03% accuracy and mitigates overfitting.","Dzongkha Next Word Prediction System  \nPrerna Chhetri* 1, Tenzin Yoezer2, Phuntsho Wangmo3, Tshewang Bomjan4.  \nInformation Technology Department, College of Science and Technology, Royal University of Bhutan, Email: [02200180.cst@rub.edu.bt](02200180.cst@rub.edu.bt)*1 , [02190165.cst@rub.edu.bt](02190165.cst@rub.edu.bt 2)[ 2](02190165.cst@rub.edu.bt 2),  \n[02200160.cst@rub.edu.bt](02200160.cst@rub.edu.bt 3)[ 3](02200160.cst@rub.edu.bt 3) , [02200170.cst@rub.edu.bt](02200170.cst@rub.edu.bt 4)[ 4](02200170.cst@rub.edu.bt 4)  \nAbstract  \nDzongkha, being the national language of Bhutan is a common and widely spoken language in the country. Official documents, scriptures and other literature products are written in Dzongkha Language in order to retain the cultural value. However, documenting Dzongkha writing is a challenging and time-consuming process, largely due to the complexity of the script, the need for multiple keystrokes per syllable, and the limited availability of efficient typing tools. An immediate system that can predict and display a list of probable words for Dzongkha Language is the solution for this problem. The project is mainly aimed to make Dzongkha typing as convenient as possible by reducing the number of keystrokes. The system can prove to be beneficial especially for Dzongkha writers. Our dataset is acquired from DCDD and has a total of 100000 sentences, 1331282 words and 28344 unique words. The data preprocessing was done by removing all the alphanumeric characters, tokenization, generating N-gram sequence and padding. Three models that were selected for training are LSTM, Bi-LSTM and GRU. The training process included fine-tuning of the model’s hyperparameters. Accuracy is the performance metrics which we used to determine the performance of each model. While training the model on a 100000 sentences dataset, GRU being lightweight and its ability to handle larger dataset performed better with 74.03% accuracy and also solved the problem of overfitting.  \nKey Words: Next word Prediction, DCDD, GRU, accuracy, overfitting.  \n1. INTRODUCTION  \nDzongkha is the national language of Bhutan. People belonging to different regions of Bhutan have their own dialect which was spoken from their ancestral times. These people use Dzongkha as a language to communicate with people of different regions and dialects as it is a common language in the country. Moreover, it is also a standard language that is spoken in the various government institutions and offices.  \nDzongkha language typing is very time consuming and also prone to a lot of errors. To address this, next word prediction systems offer a potential remedy. By suggesting Dzongkha words and reducing manual effort, these systems enhance typing speed and accuracy. This aids in preserving language authenticity. Implementing such technology can alleviate typing challenges, maintain language richness, and promote meaningful expression in Dzongkha.  \nEven though Dzongkha is being promoted through writings, messages and documentations, it is evident that Dzongkha typing is difficult (Wangchuk et Al., 2021) . Unlike typing alphabetic scripts, in Dzongkha each syllable is composed of multiple characters and because of that one would require to use more key presses to type the words which can be quite hectic and time consuming. Not to mention Dzongkha spelling is unnecessarily cumbersome.  \nKeeping all these issues in mind, there is an urgent need for a system for Dzongkha writers that can make Dzongkha typing as convenient as possible. For these reasons, this project proposes a Dzongkha Next Word Prediction System that is able to predict the next syllable, reduce keystrokes and enhance the experience of Dzongkha typing.  \n2. LITERATURE REVIEW  \nThe paper written by Wangchuk et al., mentioned that their next syllable prediction system aimed to address all the issues concerned with Dzongkha typing, including the use of several keystrokes. An LSTM neural network was incorporated","cbCaijwKJDXoyJoM","https://ap.wps.com/l/cbCaijwKJDXoyJoM","pdf",472286,1,7,"English","en",105,"# 1. INTRODUCTION\n# 2. LITERATURE REVIEW","[{\"question\":\"Why is Dzongkha typing challenging in this system’s context?\",\"answer\":\"Dzongkha typing is time-consuming and error-prone because each syllable is made of multiple characters, requiring multiple key presses, and the spelling is cumbersome.\"},{\"question\":\"What is the main goal of the Dzongkha Next Word Prediction System?\",\"answer\":\"The system aims to make Dzongkha typing more convenient by predicting the next syllable/word and reducing the number of keystrokes.\"},{\"question\":\"Which model performed best and how was performance measured?\",\"answer\":\"GRU performed best on the training dataset, reaching 74.03% accuracy. Accuracy was used as the performance metric, and GRU also helped address overfitting.\"}]",1784209811,18,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"dzongkha-next-word-prediction-system","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/dzongkha-next-word-prediction-system/86249/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is Dzongkha typing challenging in this system’s context?","Question",{"text":75,"@type":76},"Dzongkha typing is time-consuming and error-prone because each syllable is made of multiple characters, requiring multiple key presses, and the spelling is cumbersome.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What is the main goal of the Dzongkha Next Word Prediction System?",{"text":80,"@type":76},"The system aims to make Dzongkha typing more convenient by predicting the next syllable/word and reducing the number of keystrokes.",{"name":82,"@type":73,"acceptedAnswer":83},"Which model performed best and how was performance measured?",{"text":84,"@type":76},"GRU performed best on the training dataset, reaching 74.03% accuracy. Accuracy was used as the performance metric, and GRU also helped address overfitting.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,113,117,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":111,"slug":112},50,"technology",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":114,"show_sort_weight":115,"slug":116},"Healthcare",40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]