Gradofrequenza

Gradofrequenza is a metric used in data analysis and network science to describe how the frequency of a feature relates to the connectivity or involvement of that feature in a dataset or network. The concept blends two basic components: the degree of an element (grado), indicating its level of connections or category breadth, and the frequency (frequenza) of that element in the observed data.

In a network setting, for a node v, gradofrequenza can be defined as G(v) = deg(v) × f(v), where deg(v) is the degree of v and f(v) is its relative frequency in the data. Normalized variants are common, such as G_norm = (deg(v)/max_deg) × (f(v)/max_f). In text or item-collection contexts, the same idea is used by weighting a feature by both its occurrence count and its network involvement.

Applications include characterizing patterns in social networks, co-occurrence networks, linguistic corpora, and market basket data. A high gradofrequenza suggests a feature that is both well connected and prevalent, whereas a low value may indicate scarce or peripheral participation. The metric is typically used as a descriptive or exploratory tool to gain insight into structural and distributional properties of the data.

Although used in some studies, gradofrequenza is not a standard, widely adopted metric. Definitions and normalization schemes vary by domain, and the measure is often employed alongside other indicators such as centrality, degree, or frequency to support interpretation.

See also: degree, frequency, weighted degree, centrality metrics, TF-IDF, co-occurrence.