COMPARATIVE ANALYSIS OF METHODS AND MODELS FOR ESTIMATING SIZE AND EFFORT IN DESIGNING MOBILE APPLICATIONS

Mobile applications have long become an integral part of everyday life and have gained popularity among ordinary users, small companies, and large enterprises. However, this is a completely separate area of software. We can say that this is a whole separate ecosystem. Creating functional software for this ecosystem comes with new requirements, constraints, and characteristics that were not used in previous software metrics evaluation methods such as desktop programs or web applications. Mobile applications have new features, limitations and capabilities that are different from other software. Estimating the size and effort helps the software development team approximate the costs needed to more successfully develop a software product such as a mobile application. An early estimate of size and effort is necessary to identify resources that can help in the early stages of implementation a software project within budget and on time. The reliability of the assessment is also of great importance. Low reliability can lead to overestimation or underestimation, which can have bad consequences for any software development company, especially mobile applications. However, the suitability of methods and models for measuring size and efforts that are used for traditional software (PC applications and web applications) do not always provide sufficient reliability for mobile applications.The main goal of this article is to analyze the usage of various methods and models for estimating size and effort the development of mobile applications, including for the Android OS at the initial stages of their design. There is a brief description of the main and widespread methods and models that are used to estimate the size and effort, such as: Lines of Code (LOC), Function Points (FP), COCOMO and COCOMO 2, COSMIC, FiSMA, application of regression analysis for construction of mathematical models. The methods and models proposed by various authors are directly analyzed.