Question Answering Systems (QASs) have emerged as a good alternative for information seekers to retrieve precise information over the Internet. A good amount of research has been done to improve the performance of QASs across several languages, including European and Asian languages. However, Arabic, a morphologically rich Semitic language spoken by over 422 million people, has not seen similar development in the field of question answering. This article reviews the developments taking place in Arabic QASs as well as the challenges faced by researchers in developing Arabic QASs. After conducting an extensive literature survey of a number of English and Arabic QASs, this article classifies them according to several criteria. The most commonly used architecture for the development of an Arabic QAS, known as pipeline architecture, has been presented. In order to encourage and support the new researchers and scholars in conducting research in Arabic QASs, a list of techniques, tools, and computational linguistic resources, required to implement the components of the presented pipelined architecture, are described in this article in a simple and persuasive manner. Finally, the gap analysis between the research in Arabic and English QASs has been performed and accordingly, some future directions for research in Arabic QASs have been proposed.

The social media is gaining a lot of importance among businesshouses, academicians, medical practitioners, politicians, among others, due to its role in creating awareness about products, services, and socio-political views. The end users of these products, services, and views provide their feedbacks in the form of comments. An accurate determination of the sentiments of end users is crucial in designing policies and plans for products and services in future. As the processing power and storage capacities of computers have increased several folds, researchers can focus more on the accuracy of sentiment detection than consumption of computational resources. In this paper, we are applying a set of heuristics to analyse sentiments using freely available dictionary resources and open source tools. We have tested these heuristics over a large data set collected from standard sources. The experimental results are promising and opening new research directions in dictionary-based sentiment analysis.