Today Learning Management Systems (LMS) have become an integral part of learning mechanism of both learning institutes and industry. A Learning Object (LO) can be one of the atomic components of LMS. A large amount of research is conducted into identifying benchmarks for creating Learning Objects. Some of the major concerns associated with LO are size, learning outcomes, pedagogical relevance, and amount of information it delivers to learners. With the advent of knowledge enriched learning, there is a need to create Knowledge Objects (KO) as well and combine these with LOs to create Learning Knowledge Objects (LKO), which can be delivered through an LMS, so that a more holistic knowledge bank is provided to the learners. For an effective LMS, creating a high quality LKO using an algorithm that ensures the delivery of appropriate learning material to the learners is the key issue. Smaller and relevant objects can be delivered to the student using data mining approaches, thereby helping advanced learners to improve their higher order thinking skills. Use of hierarchical clustering techniques for identifying LOs based on user needs is already established. In this paper the Shared Density Approach (SDA) is used to get cohesive clusters and handle cluster of different densities. Finding similar learning objects through clustering technique reduces the domain of search. SDA not only helps with delivery of Learning Objects from a relevant cluster, but also helps in finding objects that are closer to one another but belong to a different class. Objects can be delivered based on user learning approaches, thereby have a wider usage and thus improve re-accessibility.