USE OF ARTIFICIAL INTELLIGENCE IN THE DEVELOPMENT OF ANDROID APPLICATIONS

Abstract

This paper examines the use of the WEKA machine learning library in Android applications through its porting to this platform. The main focus is on the possibilities of using WEKA in the context of emotion recognition of text on Android mobile devices. Sentiment analysis of text is an important task in the field of natural language processing, which has a wide range of applications, including social media analysis, customer feedback, and more. As part of this study, a proprietary methodology was developed for using WEKA to determine the mood of a text. This methodology is based on the use of the twitter_emotion dataset, which was divided into training and test samples for experimental comparison of different libraries. The results of the study demonstrate that the best percentage of accuracy is achieved using the WEKA library. This may be related to the relevance of the algorithm implementation and optimization of the code used in this library. In addition, a comparison of the speed of the algorithm on different Android devices was made. A difference in execution time was found, which may be due to the architectural features of mobile devices and the level of adaptation of Android for them. The obtained results will contribute to the optimal selection of the library and methods for solving the tasks of determining text sentiment on mobile devices. This work is aimed at improving the understanding of the possibilities and limitations of using machine learning in mobile applications, given the specifics of the Android platform. It also contributes to the development of new strategies and technologies for the effective use of machine learning in mobile applications. It is important to note that the use of the WEKA library in Android applications is not limited to emotion recognition of the text. WEKA can be used for a variety of machine learning tasks, including classification, regression, clustering, and more. Thus, this work can serve as a basis for further research on the use of WEKA in Android applications.