A Novel Framework for Multilingual Script Detection and Pattern Analysis in Mixed Script Queries
Open Access
- 30 September 2024
- journal article
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 43, 214-228
Abstract
A script detection system that is capable of handling several languages is becoming more necessary in today's world. The task of identifying scripts written in various languages has been substantially facilitated by the use of machine learning and deep learning, respectively. Machine learning techniques have used the Naive Bayes and Support Vector Machines (SVM) mechanism for the purpose of language detection. On the other hand, this paper reviews several unique deep-learning processes that have considered a range of methodologies, including LSTM and Bert. On the other hand, it has been shown that there is a need to improve the accuracy and the scalability often incorporated in multilingual systems. As a consequence of this, the primary focus of the present investigation is on the development of an innovative framework that is capable of recognizing scripts in a variety of languages. In addition, this technique considers pattern analysis while considering mixed script queries. A scalable, efficient, and adaptive approach has been established via study to increase the accuracy of the identification of a large number of languages. Accuracy, recall, and F1-score are some of the performance metrics that have been calculated in order to evaluate the efficacy of the multilingual script identification that has been presented. In conclusion, it has been found that the approach that was provided has supplied a solution that is both efficient and scalable for the detection of multilingual scripts.Keywords
This publication has 33 references indexed in Scilit:
- BharatBhasaNet-A Unified Framework to Identify Indian Code Mix LanguagesIEEE Access, 2024
- Deep learning and natural language processing in computation for offensive language detection in online social networks by feature selection and ensemble classification techniquesTheoretical Computer Science, 2022
- DravidianCodeMix: sentiment analysis and offensive language identification dataset for Dravidian languages in code-mixed textLanguage Resources and Evaluation, 2022
- A novel hybrid particle swarm optimization and gravitational search algorithm for multi-objective optimization of text miningApplied Soft Computing, 2020
- Emotion Detection in Hinglish(Hindi+English) Code-Mixed Social Media TextProcedia Computer Science, 2020
- Hyperparameter Optimization2019
- Machine transliteration and transliterated text retrieval: a surveySādhanā, 2018
- Hindi Roman Linguistic Framework for Retrieving Transliteration Variants using BootstrappingProcedia Computer Science, 2018
- Cross-Lingual Information Retrieval: A Dictionary-Based Query Translation Approach2017
- Discovering and understanding word level user intent in Web search queriesJournal of Web Semantics, 2015