Search

Abstract Fairness has become a critical value online, and the latest studies consider it in many problems. In recommender systems, fairness is important since the visibility of items is controlled by systems. Previous fairness-aware recommender systems assume that sufficient relationship data between users and items are available. However, it is common that new users and items are frequently introduced, and they have no relationship data yet. In this paper, we study recommendation methods to enhance fairness in a cold-start state. Fairness is more significant when the preference of a user or the popularity of an item is unknown. We propose a meta-learning-based cold-start recommendation framework called FaRM to alleviate the unfairness of recommendations. The proposed framework consists of three steps. We first propose a fairness-aware meta-path generation method to eliminate bias in sensitive attributes. In addition, we construct fairness-aware user representations through the meta-path aggregation approach. Then, we propose a novel fairness objective function and introduce a joint learning method to minimize the trade-off between relevancy and fairness. In extensive experiments with various cold-start scenarios, it is shown that FaRM is significantly superior in fairness performance while preserving relevance accuracy over previous work.

Background: The hypomania checklist-32 (HCL-32) is a widely used questionnaire developed for identifying hypomanic components in patients with a depressive episode. Measuring and screening previous hypomanic symptoms in individuals without any definite history of depressive episode would also be needed for early detection of bipolar disorders (BDs). This study aimed at testing the clinical utility of the HCL-32 for screening of BDs in the non-clinical population. Methods: Lifetime history of hypomanic symptoms was evaluated by using the HCL-32 in 220 patients with BDs and 313 non-clinical individuals. Sensitivity, specificity, and the area under the curve (AUC) of the Receiver Operating Characteristic (ROC) were evaluated for assessing the discriminatory power of the scale and its two sub-domains in screening BDs. Results: The mean HCL-32 total score was significantly higher in the Bipolar II disorder group compared to the non-clinical group (P < 0.001). Most of the items (10/12) of the irritable/risk-taking factor showed higher positive responses in patient groups. Items of active/elated factor showed mixed results. The HCL-32 total score and the active/elated factor score were not adequate for both BDs and its subgroups with AUC values of less than 0.7. The irritable/risk-taking factor score showed higher discrimination power, i.e. AUC for BDs, Bipolar I disorder, and Bipolar II disorder was 0.71, 0.67, and 0.75, respectively. Conclusions: The HCL-32 could not adequately distinguish BD patients from the non-clinical adult population. However, the current study identified items of irritable/risk-taking factor of the scale that could be useful in screening BDs in the general population. [ABSTRACT FROM AUTHOR]