||Enterprises are now a more sophisticated consumer of data which can be available in an unstructured format such as audio, images, video, and unstructured text. Data needs to be analyzed and manipulated at low cost to come up with results that gives insights for the management to enable them to take informative decisions. This is how big data gained its emerging importance in IT (information technology) world. Internal organizational customer's information and transactional data form big pool of unstructured data. Within that data lays hidden statistics that (if analyzed and sorted) would expose client's interest, trends, common behavioral patterns and much more. This enormous amount of data (Big Data) is becoming easily available and accessible due to the progressive use of technology.
The importance of this proposal comes from focusing on crucial aspect of conversation automation, by introducing a structured approach to connect big data along with NLP (Natural Language Processing) methodologies. In this proposal we are planning to build a smart chatbot for Arabic language (country of study) that teaches itself and gets smarter by fetching the patterns within enormous unstructured data pool. A chatbot is a program that performs some actions automatically within the same interface as a user. Today's chatbots are not very smart. Because it's hard to ask a chatbot the right question to get the exact answer you need. In addition, the currently implemented chatbots either do not support Arabic language (country of study) or the level of accuracy is extremely low. However, the methodology proposed here emphasizes auto learning for our chatbot where the library getting enhanced by the increasing responses over time. The more the chatbot tends to answer questions, the more its library will grow. Hence, it will become more responsive gradually. Also, Hadoop that detects new patterns in the unstructured big data pool will enhance the chatbot's engine as illustrated in Figure 1. The application areas for this project are unlimited since 80% to 90 % of
organizations data is presented in an unstructured format. Suggested application will give enterprises the opportunity to find insights within their own data and make use of it; leverage their performance and replace human interaction with a more efficient way and take their customer service to a whole new level.