Advancement of sequential mining algorithms is seen toward multiple dimensions. Alongside, the hardware technologies are also advanced in parallel. But due to the exponential growth of data, even terabytes of main memory become insufficient for handling the huge data into the memory for computing. The problem of the sequential mining algorithms lies there. Loading the whole data into the memory reduces computing ability. When computing is to be enhanced, data is to be decreased. Loading data in chunks in to the memory may skip important features of the data and loss meaningful inference rules. Reading dataset in multiple passes increases computing overhead. So finding an algorithm that manages an optimum tradeoff between the memory consumption and computing efficiency is really a great challenge for the researchers today. Use of proper data structures in algorithms with database scan reduction techniques can elevate the sequence mining process up to manifolds. In this paper, the AprioriAll sequence mining algorithm and its variants are implemented by using Hash and TRIE structures with database scan reduction techniques. We consider Nifty 50 index data for studying the algorithms. A thorough study is presented to compare the generated results against different performance parameters like support, confidence, dataset size, scan reductions, and space requirements. It is observed that the AprioriAll with TRIE structures outperforms the AprioriAll with Hashing in respect of processing time whereas AprioriAll with Hashing is a better choice to make in respect of memory consumptions.