Author(s)

P. Mariani & G. Ventre

Abstract

Methodology represents an original application of text mining aimed at utilising Broadbent’s model in the pharmaceutical ambit. This approach offers a quantitative measure of the message scatter. From communication from the emitter, the receiver absorbs partly or wholly the communication, but can adapt the contents by its experience or by \“communication filters”. Measurement can be performed by asking the receiver to describe the communication contents by means of verbalisation of the communication itself. The use of multiple correspondence analysis and cluster analysis has permitted a quantification of the main dimension beneath the data and a division in patterns on the basis of different linguistic models to be obtained. A test has been performed on a sample of 51 articles extracted from \“NOOS”, a magazine of psychiatric argument. Four articles were chosen like the model and compared with the rest of the sample, verifying textual correlation and related with the four articles model. In this way a simulation of the communication flow has been made from the Company (Medical Department) to the physicians, passing through different staff of the firm, particularly REP, and it has been possible to measure the quota of correct non-scattered communication between companies and physicians. Keywords: text mining, communication scatter. 1 Introduction The number of physicians enrolled in the Italian Physician Roll are 347.759 (Federazione Nazionale degli Ordini dei medici chirurghi e degli Odontoiatri,

Keywords

text mining, communication scatter.