Accounting Mining Data Using Neural Networks (Case study)

Abstract

Business organizations have faced many challenges in recent times, most important of which is information technology, because it is widely spread and easy to use. Its use has led to an increase in the amount of data that business organizations deal with an unprecedented manner. The amount of data available through the internet is a problem that many parties seek to find solutions for. Why is it available there in this huge amount randomly? Many expectations have revealed that in 2017, there will be devices connected to the internet estimated at three times the population of the Earth, and in 2015 more than one and a half billion gigabytes of data was transferred every minute globally. Thus, the so-called data mining emerged as a technique that aims at extracting knowledge from huge amounts of data, based on mathematical algorithms, which are the basis for data mining. They are derived from many sciences such as statistics, mathematics, logic, learning science, artificial intelligence, expert systems, form-recognition science, and other sciences, which are considered smart and non-traditional. The problem of the research states that the steady increase in the amount of data, as well as the emergence of many current areas that require different data due to the contemporary environment of business organizations today, make information systems unable to meet the needs of these current organizations, and this applies exactly to accounting information systems as they are the main system in business organizations today. These systems have been designed to meet specific needs that make it impossible today to meet the different needs according to the contemporary environment of business organizations today, as well as failing to deal with the amount of data generated by the information technologies. The research proposes two main hypotheses. First, the adoption of accounting data mining leads to providing data that the accounting information system was unable to provide before, as well as to shortening the time and effort required to obtain it. Second, the adoption of accounting exploration of data enables the adoption of artificial intelligence methods in processing such data to provide useful information to rationalize decisions. The research leads to a number of conclusions, including that the steady increase in the amount of data in general, and the accounting data in particular, makes dealing with traditional frameworks a very difficult issue and leads to loss of time and effort during extracting information. In addition, the emergence of many current variables as a result of changes in the work environment requires the presence of technical tools, which have enough flexibility to deal with them. Moreover, data mining tools have the ability to derive relationships based on their existing databases that were not available before. The research presents a number of recommendations, most important of which is the need to adopt the model presented by the research, i.e., Multilayer Perception, a network that exists within the (SPSS) program, which allows the possibility to use this network easily in rationalizing the decision to choose implemented projects in the provincial councils.