Mining Frequent Itemsets has become an integral component of research lately, mainly because it can be applied to a variety of areas, including clustering of documents for information retrieval, inventory management, association rule mining, and outlier detection. Existing techniques that extract "frequent items" do not take care to address redundancy. An itemset is said to be redundant if their support can be derived from other existing itemsets using standard deduction techniques. The presence of such itemsets in the collection was found to increase the size of the collection of the mined frequent items. This in turn affects the memory and run time of the mining process. To overcome these drawbacks, this paper employs a list-based approach that generates only "Closed and Non-Derivable Itemsets." This is a collection of items 'I' such that 'I' has no superset with the same support as that of itself, and I belongs to the collection of those items whose lowest upper bound and highest lower bound values are not the same. With the help of experiments, it has been proved that the proposed approach addresses redundancy better by generating a compressed set of non-redundant frequent items.