Indian Journal of Science and Technology

Abstract

Pattern matching technology not only utilizes artificial intelligence and cognitive science to process data, it also uses intelligent systems to effectively present and handle data, making it widely used in diverse areas such as finance, manufacturing, sports, and the service sectors. This paper proposes an algorithm that makes it easier for users to search for information that is relatively unstructured compared to existing multimedia data, by converting it into images. Rather than using just one piece of information to gather search results, symbols and color information can be designated to data to allow optimum search results. Under the assumption that a complete database of the world exists, the search results as well as the efficiency of the system would depend on the accuracy of the image drawn for the search. However, more important than the accuracy of the results is the ability of the algorithm to comprehend the user’s intent and display search results accordingly. Moreover, seeing how the efficiency of the system might depend on the way the individual algorithms are combined, improvements on the search image generation module and further studies would allow people to identify methods that can improve the individual components of the whole algorithm as well as find better combinations of individual algorithms. The algorithm proposed in this study has been seen to lead to an improvement of individual matching algorithms. It is of utmost importance that an algorithm with a human-like recognition system is developed in order to create a system based on pattern recognition.
Keywords: Big Data, Cognitive System, Emoticon, Pattern Matching, Pattern Recognition, Sensing