Authors : T. Chellatamilan and R.M. Suresh
Abstract: In current scenario, retrieving appropriate data from vast data repositories in superlative method is a confrontation. Since, there is a greater competency on the number of pages indexed and the retrieval speed of relevant documents in Information Retrieval (IR) through which the people access information from distinct data stores, disparate methodologies have been developed. In order to reduce noise and to acquire more precise results, researchers proposed an algorithm called Affinity based Nominal Language Model (NLM). The LM basically composes some probability measures which in turn bestows rank to the retrieved documents that afford an essence for the adequate information retrieval process. The methodology comprises the customs of IR such as query expansion, clustering and document ranking and the predominant Nominal Language Model which is working with the description of part of speech of the language that describes the features, combined nouns and adjectives. The NLM in the proposal deals with the affinity rate calculation that is based on the affinity against the document over user concern, combined with probabilistic measurements to find the occurrence rate of a particular term of query within the document. This begets the algorithm to produce optimal results for the applied query with high accuracy rate, less processing time and the minimal use of memory in an adept manner.
T. Chellatamilan and R.M. Suresh, 2013. Affinity Based Nominal Language Model (NLM): A Dynamite Information Retrieval Approach. International Journal of Soft Computing, 8: 72-81.