Abstract In natural language processing, identifying semantic similarity is a major challenge for which several measures have been introduced and tested. There are four types of semantic measures which includes the following. First are the Ontology Based measures that depend on the closeness of the concepts in the taxonomy. Ontology-based measure detects the similarity in terms of the path linking the... In natural language processing, identifying semantic similarity is a major challenge for which several measures have been introduced and tested. There are four types of semantic measures which includes the following. First are the Ontology Based measures that depend on the closeness of the concepts in the taxonomy. Ontology-based measure detects the similarity in terms of the path linking the concepts and position of the concept in the hierarchy. The second one is related to the Information Content of concepts is considered to find the semantic similarity. The third and fourth includes the Featurebased similarity and Hybrid similarity measures. Semantic similarity measurements have wider impact in other areas such as data mining, computational intelligence, linguistics, information retrieval systems and so on. Whatever the measure is used, arriving at the quality is a prominent question. For identifying the effectiveness, the systems like the Psycholinguistic evaluation is used to justify the similarity measure quality. The comparison of the proposed semantic similarity measure values is being carried out with the expert opinion. The values are statistically tested suing Pearson Correlation Coefficient is used to test the quality of the similarity measure. When the correlation between the computational method value and the human assessment values. This work described the proposed ASC semantic similarity measure and its Psycholinguistic evaluation versus the opinion of the experts. Read More Read Less

Abstract A comparison of the developed complex Hadamard transform with the well-known unitary (orthogonal) transforms of Karhunen-Loeve, Fourier, Walsh-Hadamard and discrete cosine transform is presented. The comparison is made on the base of minimization of mean-squared error of reconstructed transform coefficients for two test images.... A comparison of the developed complex Hadamard transform with the well-known unitary (orthogonal) transforms of Karhunen-Loeve, Fourier, Walsh-Hadamard and discrete cosine transform is presented. The comparison is made on the base of minimization of mean-squared error of reconstructed transform coefficients for two test images. Read More Read Less

Abstract In this work, we have presented the optimization of the parameters of two-parameter Keys kernel and did the Optimization using the spectral domain. The optimization criterion is the condition that the amplitude characteristic is a good approximation of the amplitude characteristic of the ideal interpolation kernel of the form sinx/x. Further, the work presented the experimentally determined optimal value of... In this work, we have presented the optimization of the parameters of two-parameter Keys kernel and did the Optimization using the spectral domain. The optimization criterion is the condition that the amplitude characteristic is a good approximation of the amplitude characteristic of the ideal interpolation kernel of the form sinx/x. Further, the work presented the experimentally determined optimal value of the kernel parameters using signals. Read More Read Less