No abstract

No abstract

Recommender systems have become very important for many online activities, such as watching movies, shopping for products, and connecting with friends on social networks. User behavioral analysis and user feedback (both explicit and implicit) modeling are crucial for the improvement of any online recommender system. Widely adopted recommender systems at LinkedIn such as "People You May Know" and "Endorsements" are evolving by analyzing user behaviors on impressed recommendation items.In this paper, we address modeling impression discounting of recommended items, that is, how to model user's no-action feedback on impressed recommended items. The main contributions of this paper include (1) large-scale analysis of impression data from LinkedIn and KDD Cup; (2) novel anti-noise regression techniques, and its application to learn four different impression discounting functions including linear decay, inverse decay, exponential decay, and quadratic decay; (3) applying these impression discounting functions to LinkedIn's "People You May Know" and "Endorsements" recommender systems.

13-cis-Retinoic acid (13-cis-RA), also known as isotretinoin, is commonly used in the management of severe acne. Its clinical efficacy in oncology has also been documented. As a vitamin A derivative, it is not soluble in water. This solubility barrier not only affects its oral absorption but also makes parenteral delivery difficult. Recently, water-soluble formulations of 13-cis-RA have been attempted with 2-hydroxypropyl-beta-cyclodextrin (HP-beta-CD) and randomly methylated-beta-cyclodextrin (RM-beta-CD). In this study, the pharmacokinetic profiles of these two formulations were assessed in Sprague-Dawley rats after single intravenous or oral administration. We found that 13-cis-RA was eliminated from the body through a dose-independent process after intravenous injection of either sodium salt or the HP-beta-CD formulation within the tested dosage range (2.0-7.5mg/kg). Furthermore, HP-beta-CD did not alter the kinetic profile of 13-cis-RA after intravenous administration in comparison with 13-cis-RA sodium salt. We also found that RM-beta-CD dramatically enhanced the oral absorption of 13-cis-RA. At 10.0mg/kg, the bioavailability of 13-cis-RA formulated with RM-beta-CD was about three-fold higher than that of the control (13-cis-RA suspended in 0.5% carboxymethylcellulose (CMC)). Similarly, the oral absorption of 13-cis-RA was not saturated within our tested range (2.5-10.0mg/kg) and the bioavailability remained unchanged. These results demonstrated that HP-beta-CD and RM-beta-CD were suitable excipients for the delivery of 13-cis-RA.

Quasi-cliques are an elegant way to model dense subgraphs, with each node adjacent to at least a fraction λ ∈ (0, 1] of other nodes in the subgraph. In this paper, we focus on a new graph mining problem, the query-driven maximum quasi-clique (QMQ) search, which aims to find the largest λ-quasi-clique containing a given query node set S. This problem has many applications and is proved to be NP-Hard and inapproximable. To solve the problem efficiently in practice, we propose the notion of core tree to organize dense subgraphs recursively, which reduces the search space and effectively helps find the solution within a few tree traversals. To optimize a solution to a better solution, we introduce three refinement operations: Add, Remove and Swap. We propose two iterative maximization algorithms, DIM and SUM, to approach QMQ by deterministic and stochastic means respectively. With extensive experiments on three real datasets, we demonstrate that our algorithms significantly outperform several baselines in running time and the quality.