Abstract: This disclosure pertains to methods and systems for predicting demographics and psychographics of any application object (user or item), and matching them to other objects based on an aggregated similarity scores derived from location and unstructured textual data. In some embodiments, vector-space representation may be used for matching, with demographic weights dynamically computed, for example, based on census data distribution information. Psychographic similarity may be computed, in some embodiments, on the basis of both non-classical semantic relatedness—finding implicit connections between terms in a knowledge source, and categorical information embedded in the knowledge source, preferably a large corpus such as Wikipedia.

Abstract: This disclosure pertains to methods and systems for predicting demographics and psychographics of any application object (user or item), and matching them to other objects based on an aggregated similarity scores derived from location and unstructured textual data. In some embodiments, vector-space representation may be used for matching, with demographic weights dynamically computed, for example, based on census data distribution information. Psychographic similarity may be computed, in some embodiments, on the basis of both non-classical semantic relatedness—finding implicit connections between terms in a knowledge source, and categorical information embedded in the knowledge source, preferably a large corpus such as Wikipedia.