د. سلام حسن محيسن العكبي

This work aims to design and implement a rule text-mining based algorithm to analyse the headlines automatically and implement the analysing program called News Analysis Program (NAP), which is based on this algorithm and may help to support the short-term investment decision makers as a part of the Decision Support System. A manual analysis of 1133 headlines was done, which was used for selecting the keywords for bag of words and further training and verification. The linguistic dictionary Harvard IV Psycho Social and the software Wordsmith 4 were used as a part of the bag of words used in this algorithm. Alarabia.net and Reuters.com news are treated as a source of media noise that has an influence on the value of stock quoted on the stock market. The correspondence ratio between the manual and automatic analysis is 88.79% for the pattern that reflected the effect of news on the bank sector's stocks throughout October, November and December 2012. News Analysis Program is implemented in the Python programming environment .

The main goal of this paper is to present what the Islamic banks (the type of banks governs by Islamic law that governs every aspect of Muslims’ life “Shariah”) instruments are and why it is so important for the community. According to Shariah the payment or acceptance of interest charges (Riba) is prohibited in Islamic banking transactions. This paper is tried to describe the main features of the Islamic banking transactions and to identify the differences between Islamic and conventional banks. Also, this paper presents a survey of the historical evolution of the Islamic banking system in Muslim countries, starting with the first Islamic bank, early in the 1960’s, Nasser Bank in Egypt. In conclusion, the Islamic banking system is a rapid growth one and the switch to Islamic banking will be a difficult task. Due to the fact, that in the last decade, the Islamic banking made its presence in all over the world markets not just in Islamic countries

This work aims to use vector measure construction method (VMCM) in classification Gulf Cooperation Council (GCC) stock markets ratios movement in term of its response to the media news indicator. VMCM is used for classified dynamic set of data such as showing dynamic stock market changes related to media influence as in this article. Two types of data are used in this article throughout October, November, and December 2012. The first type is the news indicator Good News Indicator (GNI), Bad News Indicator (BNI) and, Neutral News Indicator (NNI) as a media expansiveness measurement. The headlines collected from Alarabia.net and Reuters.com. The second kind of data is the stock market ratios of banking and energy companies on the GCC stock markets. This work shows the dynamic changes of the quieted stock to the news that used here as a noise source

This paper presents the effect of news on the oil and gas sectors onthe Kuwait Stock Exchange (KSE)for 92 days, and 63 trading days, throughout October, November, and December of 2012, choosing the optimum period with the best reaction to the news. Using classical methods,correlation and at-testwere used to measure and verify significance. The resultsshowed thatgood news indicator and bad news indicator indices affected theoretical predictions and led to a significant improvement,compared to the efforts ofprevious researchers in this domain. There were profound implications for future studies of the behavior of Kuwait's stock markets.

This paper presents a comparison of distance measure methods of K-means clustering method on the example of the bank sector companies of Golf Cooperation Council Countries (GCC) stock markets. These companies are treated as a pattern for the effect of news on the bank sector’s stocks of the GCC Stock markets for 92 days, and 63 trading days, throughout October, November, and December of 2012. Analysis allows choosing the optimum period reflected the best reaction to the news. Correlation coefficients and t-statistics for GNI and BNI and financial factors like PER, PBV, DY and rate of return were used as diagnostic variables for Distance measure methods

This paper presents the effect of news on AbuDhabi Security Exchange (ADX)for the 64days and 40 trading days throughoutNovember and December of 2011. Using classical methodsthe correlation and regression were measured. ResultsshowthatGNI and BNI indices have affected theoretical predictions and significant improvement over previous efforts by pervious researcher in this domain. The work presented here has profound implications for future studies of UAE’s stock markets behavior.

This work aims to design a proposed decision support system (DSS) for helping investors in making investment decision by using rule text-mining based algorithm to analyze news headline and implement analyzing pro-gram based on a manual analyzed headlines. The news analysis program (NAP) was used as an important stage in making investment decision on sample of the Gulf Cooperation Council (GCC) stock markets using Alarabia.net and Reuters.com which treated as a source of media noise that has an influence on the value of stock quoted stock market. The second kind of data that proposed to use in this system is the financial data of GCC stock market. The resulted data can be used in further steps to make better understanding of stock market companies behavior such as the statistical, data mining calculation for choosing the best period of time that give the best reaction of stock market ratios to the news indicators and using the vector measure construction method (VMCM) for classifying companies according to their response to the news.

Abstract— This paper introduces an experimental study of the heart disease datasets characteristics impact on the performance of classification algorithms in the aim of identifying the best algorithm for each dataset under its characteristics. The performance of five machine learning algorithms (logistic regression (LR), K-Nearest Neighbor (KNN), Decision tree (DT), Random Forest (RF), and support vector machine (SVM)), single layer neural network (ANN), and deep neural network (DNN), has been evaluated using five heart disease datasets under four data complexity measurement: number of samples (dataset size), number of features (dimension of dataset), Data sparsity measures, and correlation of features. All datasets have been processed and normalized then the mutual information-based feature selection method was used to solve the overfitting problem. The results show that in general, the machine learning especially the Random Forest algorithm achieves high classification accuracy than deep learning network. In other hand, the high sparsity and less mutual information of dataset has large impact on degradation of the performance of classification algorithms than other characteristics of data.