prediction of ICT equipment new, reused, or refurbished

The growing volumes of end-of-life (EOL) and near end-of-life ICT equipment around the globe is becoming a matter of concern, particularly as ICT equipment is characterized by high demand and a relatively short lifespan. Typical lifespans can vary and will depend on how well your assets were maintained throughout their lifetime [13].

During the five-year period of GTP I (2009/10-2014/15), access to ICT networks expanded significantly. The number of mobile subscribers increased from 6 million to 39 million. Similarly, the percentage of the rural population within 5 kilometers radius to telephone service increased from 62 per cent to 96 per cent. The number of Internet usage rose from fewer than a million to almost more than 4 million. Growth was also experienced in networks connecting government and social institutions such as secondary schools and hospitals [12].

ICT equipment should match with role or mission of the organization. In SNNPR public sector there is less effective usage of ICT equipment [3]. The research is proposed to model that predicts the status of ICT equipment with in the regional public sector using datamining techniques. In the region, large quantity of ICT equipment is purchased in every year but by simple defects equipment are thrown into the organizations warehouse.

Data mining has been defined as the non-trivial extraction of earlier unknown, implied and potentially useful information from data. It is the science of mining useful information from huge databases. Data mining is one task in the procedure of knowledge discovery from the database. There are two major data mining tasks: predictive and descriptive modeling. Predictive modeling is predicting the outcome of a database, is one of the most motivating and challenging tasks where to develop data mining requests. Descriptive DM tasks is describing the general properties of the data in the database, that is, produce new, nontrivial information based on the available data set (Gupta et al., 2011).

Predictive modeling, a tool used in predictive analytics, refers to the process of using mathematical and computational methods to develop predictive models that examine current and historical datasets for underlying patterns and calculate the probability of an outcome. The predictive modeling process starts with data collection, then a statistical model is formulated, predictions are made, and the model is revised as new data becomes available.

Predictive modeling is generally categorized as either parametric or nonparametric models. Within these two camps are several different varieties of predictive analytics models, including Ordinary Least Squares, Generalized Linear Models, Logistic Regression, Random Forests, Decision Trees, Neural Networks, and Multivariate Adaptive Regression Splines. Therefore, by using the model technicians may easily predict the status of ICT equipment whether it could be reused, recycled or disposed. The research will therefore aim to construct a predictive model by using local dataset for the status of ICT equipment.

2. Problem statement

The country has made equally impressive strides in spreading ICTs across the government, education and health sectors and developing the computer software and ICT manufacturing sectors.

Nevertheless, the adoption of ICT in Ethiopia remains low in comparison with regional and global peers. The major constraints that are preventing Ethiopia from achieving its full digital potential include: [12]

The need to update legal and regulatory frameworks to deal with new technological advances and, with recent infrastructure growth, to enhance monitoring of service affordability and quality.

The need to increase the usage of broadband service in the country.

Constraints in take up of ICT services in rural areas by citizens, businesses and health and educational institutions due to awareness issues.

The need to avail affordable ICT devices and services for some elements of the population.

The organization have no clear status of the ICT equipment

Mass of ICT equipment are in organization’s warehouse by simple defect.

3. Research questions

This paper attempts to answer the following questions:

• How to develop a predictive model for predicting the status of ICT equipment in public offices of SNNPR?
• Which attributes are more important to predict the status of ICT equipment?
• Which classification algorithm can be more suitable for the purpose of predicting status of ICT equipment?

4. Objective
   • General Objective

The general objective of the research is to predict the status of ICT equipment using data mining technique in case of Southern Region public sectors.

• Specific Objectives

1. To understand the area via an extensive literature review.
2. To prepare data sources, process the dataset.
3. To identify the important attributes used to determine the s status of ICT equipment used in public sector.
4. To find out classification algorithms that more suitable to build predictive model for the status ICT equipment.
5. To develop a predictive model for predicting the status of ICT equipment.
6. To evaluate the performance of the models.
7. Literature review

In modern society ICT is ever-present, with over three billion people having access to the Internet. With approximately 8 out of 10 Internet users owning a smartphone, information and data are increasing by leaps and bounds. The Ethiopian ICT sector grew substantially over the last decade but it remains underdeveloped compared to its African peers [1].

Ethiopia’s ICT infrastructure has seen significant improvements in the past few years. Mobile telecommunications alone grew from a mere 1 million subscribers in 2007 to 39 million in 2014/15 [12].

6. Methodology

This thesis will use a Hybrid data mining model which is a six-step knowledge discovery process model. Due to the nature of the problem and attributes in the dataset, classification mining task will be selected to build the predictive model.

• Understanding the problem

At this stage to understanding the problem, an extensive literature review will perform to study the existing problems in SNNPR ICT sector to determine the status ICT equipment and done consultation with IT experts to have an in-depth insight into the problem.

• Understanding the data

The main source of the data to be used to undertake this research will be taken from SNNPR ICT sector. The main reason to select this sector is there are large amount of ICT equipment to be maintained from different sectors. The five years data from (2009– 2014 E.C) will be taken from SNNPR ICT sectors. The total data will obtain from SNNPR ICT sector is around 3000 and 10 attributes. Hence, the availability of the ICT sector data in Excel file format and then converts to CSV and in to Attribute Relation File Format (ARFF) file format.

• Preparation of the data

Highly preparation of data will need to get accurate prediction result. The data will be pre-processed by employing data cleaning and data selection techniques. Data cleaning refers to the pre-processing of data to remove or reduce noise, binning, and the treatment of missing values.

• Modeling

Weka will be selected as DM tool, since the whole suite of Weka is written in java, so it can be run on any platform. The mining algorithms: J48 and Naive Bayes will use in all experiments due to their high score in classification algorism. The model formed by two experiments: 10 cross validation and 75/25 percentage split.

• Evaluation and use of knowledge

The confusion matrix and ROC curve will use to evaluate the accuracy and performance of the models built with the three algorithms. From the confusion matrix, the Specificity, Sensitivity, Accuracy Rate and Error rate will calculate.

7. Scope and delimitation of the study
   • Scope

The scope of the study includes fifty public sector offices within the Southern Nation, Nationalities and People’s Region. And focuses on ICT hardware equipment.

• Delimitation

Time and financial constraints to accommodate all zone and weredas

8. Significance of study

The study is significant from the following perspectives.

Firstly, it provides information about the status of ICT equipment in the SNNPR state public sector organizations.

Secondly, it creates awareness within the regional government and public-sector.

Thirdly it replaces manual prediction and enables officials further decision based on the result.

9. Tentative work plan

10. Budget