Abstract
In the present paper we attempted to generate a parametric model for word frequencies. In order to make this relation applicable, we arranged word lengths in accordance with their normalized frequencies. The pattern of occurrence of words containing different numbers of letters has been investigated on the basis of their Zipf's order and by applying power law for Zipf's order and frequencies. The applicability of the generated mathematical model for word length frequencies was verified for different texts. We also resolved the problem of establishing a relationship between word frequencies of higher Zipf's order with text length.
Acknowledgement
The authors are grateful to the Council of Scientific and Industrial Research (CSIR), New Delhi, for providing financial assistance to carry out this research into this interesting field of applied mathematics in the form of a senior research fellowship to the first author (CSIR grant number: 09/428(0062)/2006-EMR-1).