Collaborative filtering is a fundamental building block in many recommender systems. While most of the existing collaborative filtering methods focus on explicit, multi-class settings (e.g., 1-5 stars in movie recommendation), many real-world applications actually belong to the one-class setting where user feedback is implicitly expressed (e.g., views in news recommendation and video recommendation). The main challenges in such one-class setting include the ambiguity of the unobserved examples and the sparseness of existing positive examples.In this paper, we propose a dual-regularized model for one-class collaborative filtering. In particular, we address the ambiguity challenge by integrating two state-of-the-art one-class collaborative filtering methods to enjoy the best of both worlds. We tackle the sparseness challenge by exploiting the side information from both users and items. Moreover, we propose efficient algorithms to solve the proposed model. Extensive experimental evaluations on two real data sets demonstrate that our method achieves significant improvement over the state-of-the-art methods. Overall, the proposed method leads to 7.9% -21.1% improvement over its best known competitors in terms of prediction accuracy, while enjoying the linear scalability.

The hydrolysis reaction of N,N-dimethyl-N'-(2-oxo-1, 2-dihydro-pyrimidinyl)formamidine (DMPFA), a model compound of the antivirus drug amidine-3TC (3TC = 2', 3'-dideoxy-3'-thiacytidine), is investigated by the hybrid density functional theory B3LYP/6-31+G (d,p) method. The hydrolysis reaction of the title compound is predicted to undergo via two pathways, each of which is a stepwise process. Path A is the addition of H2O to the C=N double bond in the amidine group to form a tetrahedral structure in its first step, and then the transfer of the H atom of hydroxyl leads to the corresponding products via four possible channels. Path B simultaneously involves the nucleophilic attack of H2O to the C atom of the C=N bond and the proton transfer to the N atom of amino group leading to the cleavage of the C-N single bond in the amidine group. The results indicate that path A is more favorable than path B in the gas phase. Moreover, to simulate the title reaction in aqueous solution, water-assisted mechanism and the cluster-continuum model, based on the SCRF/CPCM model, are taken into account in our work. The results indicate that it is rational for two water molecules served as a bridge to assist in the first step of path A and that cytosine rather than the cytosine-substituted formamide should be released from the tetrahedral intermediate via s six-membered cycle transition state (channel 2). Our calculations exhibit that the process toward the tetrahedral intermediate is the rate-determining step both in the gas phase and in aqueous solution.

Density functional theory (B3LYP) and ab initio (MP2) methods with the 6-31G(d,p) basis set are used to study the mechanisms for the hydrolysis of N-(2-oxo-1,2-dihydro-pyrimidinyl) formamide (PFA) in the gas phase. The direct and the water-assisted hydrolysis processes are considered, involving one and two water molecules, respectively. Three different pathways are explored in each case. In the first pathway, the O atom of water first attacks at the C atom of amide while one H atom of water transfers toward the oxygen of amide, leading to an intermediate of tetrahedral coordinated carbon with two OH groups. In the subsequent step, the hydroxyl H atom transfers to the N atom of pyrimidine ring and the C-N covalent bond of amide dissociates simultaneously. In the second path, the O and one H of water attack at the C of amide and the N of pyrimidine ring, respectively, while the C-N bond of amide dissociates. In the third path, three processes occur simultaneously: the O of water attacks at the C of amide, one H atom attacks at the N of amide, and the C-N bond of amide is broken. It is shown that the second pathway is favored for the direct hydrolysis while the first pathway is favored for the water-assisted hydrolysis. It is also shown that the water-assisted hydrolysis is slightly more favorable than the direct hydrolysis. Moreover, solvent effects on five pathways are evaluated with Monte Carlo simulation (MC) and free energy perturbation methods. It is shown that the solvent water slightly reduces the energy barrier in each pathway. The first pathway in the water-assisted hydrolysis remains the most favorable when the solvent effects of bulk water are taken into account.

No abstract