Abstract: In a distributed method of data gathering, connotative meanings of terms are selected from predefined emotional descriptors organized into a plurality of emotional categories. Respective judges select zero or one emotional descriptors from a given category and rate the term for intensity. Each term is evaluated by a plurality of judges for each emotional category. The results are processed to assure that the selected emotional descriptors were not selected by chance. Retained associations become part of a database of terms, including connotative meanings and related intensity. A self-calibration method is used during intensity evaluation to assure uniformity among intensity ratings. The judges select among anchor terms for each intensity rating in a given emotional category, picking the anchor terms which most accurately reflects the rating level being checked. The completed data base is accessed through a computer interface to implement an electronic language reference tool.

Abstract: In a distributed method of data gathering, connotative meanings of terms are selected from predefined emotional descriptors organized into a plurality of emotional categories. Respective judges select zero or one emotional descriptors from a given category and rate the term for intensity. Each term is evaluated by a plurality of judges for each emotional category. The results are processed to assure that the selected emotional descriptors were not selected by chance. Anomalous results are purged. Statistical analysis is performed to assure that the selected emotional descriptors were not selected by chance. Retained associations become part of the database for the corresponding record. There may be 0, 1 or more connotative associations for any given record. The completed data base is accessed through a computer interface to implement an electronic language reference tool.