Received: June 7, 2021. Revised: August 13, 2021. 596 International Journal of Intelligent Engineering and Systems, Vol.14, No.5, 2021 DOI: 10.22266/ijies2021.1031.52 Minimum Search-time Algorithm for Image Retrieval in Cloud Computing Abubakar Usman Othman 1 * Aisha Yahaya Umar 2 Maryam Maishanu 1 Hauwa Abubakar 3 Boukari Souley 1 Abdulsalam Ya’u Gital 1 1 Faculty of Science, Department of Mathematical Sciences, Abubakar Tafawa Balewa University Bauchi, Nigeria 2 Department of Computer Science, Gombe State University, Nigeria 3 Department of Computer Science, Umar Suleiman College of Education, Gashua, Nigeria * Corresponding author’s Email: othman80s@yahoo.com Abstract: Finding approximate nearest neighbour (ANN) is essential in huge database for efficient similarity search to return the nearest neighbour of a given query. Many hashing algorithms have been designed to improve retrieval accuracy and storage requirements of data in a large-scale database through long code word which increases the time complexity in loading data into memory, but do not consider the search time which is an important parameter in the field of information retrieval and pattern recognition. To address the aforementioned problem, this research therefore proposes an improved search time algorithm for improving the retrieval time of data from a database in cloud computing environment by optimising both the search accuracy and search time simultaneously. We improved the minimum search time by the use of balance partitioning algorithm for the even distribution of data points into hash buckets to minimise search time, and similarity preserving algorithm for search accuracy were designed for fast and accurate retrieval of data in a database. An extensive experiment conducted on a cloud simulator and the result obtained when the code length is 8, 96 bits, the retrieval time for the proposed system is 0.030sec, 0.260sec, and that of Density Sensitive Hashing is 0.040, 0.400sec. Therefore, the retrieval difference is 0.010sec and 0.140sec. Also, the result obtained for the rest of the code lengths of 16, 32 and 64 show that the improved minimum search time algorithm outperforms the compared techniques in terms of the velocity of big data retrieval. Keywords: Balance partitioning, CloudSim, Cloud computing, Data, Hashing, Information retrieval, Time. 1. Introduction Cloud computing is a web-based application that provides a shared pool of resources. The advance in mobile technology have allowed mobile devices such as smartphones and tablets to be used in a variety of different applications [1]. The availability of internet such as with the use of the wide spread broadband Internet access [2, 3], coupled with these hand held devices (mobile devices), resulted to the easy collection of digital information in form of structured and unstructured [4] data, had contributed to the availability of large volumes of data known as big data. Tremendous amount of data are generated every day in Manufacturing, Business, Financial Services, Science sectors and human personal lives. Adequate and proper processing of these data is required to open new discoveries and knowledge concerning markets, societies and human environment [5-7]. As unstructured data contributed to the availability of big data, they need to be structuralised for its effective understanding and processing through some optimised techniques used for extracting information. These information extracting techniques have been vastly used to extract meaningful information from raw or unstructured data [8]. Data has greatly changed and influenced researches in sciences. The Sloan digital sky survey is used by astronomers nowadays as a pool of resources which serve as a database [9, 10]. Biological data and experimental data are stored in a public storage facility and databases are created such that other biologists and scientists can make use of these generated biological and scientific data. Many hashing based indexing techniques were also proposed to overcome the growing volume and