A Maximum Entropy Approach for Collaborative Filtering

Collaborative filtering (CF) involves predicting the preferences of a user for a set of items given partial knowledge of the user's preferences for other items, while leveraging a database of profiles for other users. CF has applications e.g. in predicting Web sites a person will visit and in recommending products. Fundamentally, CF is a pattern recognition task, but a formidable one, often involving a huge feature space, a large data set, and many missing features. Even more daunting is the fact that a CF inference engine must be capable of predicting any (user-selected) items, given any available set of partial knowledge on the user's other preferences. In other words, the model must be designed to solve any of a huge (combinatoric) set of possible inference tasks. CF techniques include memory-based, classification-based, and statistical modelling approaches. Among these, modelling approaches scale best with large data sets and are the most adept at handling missing features. The disadvantage of these methods lies in the statistical assumptions (e.g. feature independence), which may be unjustified. To address this shortcoming we propose a new model-based CF method, based on the maximum entropy principle. For the MS Web application, the new method is demonstrated to outperform a number of CF approaches, including naive Bayes and latent variable (cluster) models, support vector machines (SVMs), and the (Pearson) correlation method.     

The prospects of non-volatile phase-change RAM

Phase-change read-and-write memory (PRAM) is a promising memory that can solve the problems of conventional memory—scalability, read/write speed, and reliability. We will review the opportunities and technical challenges of PRAM. The most important challenge of PRAM is the reduction of the writing current. Various approaches to reduce the writing current will be reviewed and the prospects of PRAM are discussed.     

Butyltin Species and Inorganic Tin in Water and Sediment of the Detroit and St. Clair Rivers

Water and sediment samples from 29 locations in the Detroit and St. Clair rivers were analyzed for the highly toxic tri-n-butyltin (Bu₃Sn⁺) species and for the less toxic di-n-butyltin (Bu₂Sn²⁺) and n-butyltin (BuSn³⁺) species and inorganic tin. In general, locations sampled in the St. Clair River were less contaminated with butyltin species than those in the Detroit River. Inorganic tin and BuSn³⁺ were detected in over 90% of all subsurface water samples, while Bu₂Sn²⁺ and Bu₃Sn⁺ were detected in 45 and 28% of the same samples, respectively. The highest concentration ofBu₃Sn⁺ in subsurface water, 5.9 × 10^?10 mol Sn/L, was at the mouth of the Ecorse River, a tributary of the Detroit River. The three butyltin species and inorganic tin were also detected in 23–46% of all sediment samples. The highest concentrations of Bu₃Sn⁺ in sediment were found close to the mouths of the River Rouge, another tributary of the Detroit River, and the Ecorse River, and were 6.2 × I0^–7 and 1.7 × I0^–7 mol Sn/kg dry weight, respectively, for the top 2 cm of sediment.     

The morphology of rigid-rod polymers and molecular composites resulting from coagulation processing

The effects of processing conditions on the morphology of molecular composite films are examined by optical and electron microscopy. During coagulation processing from solutions in methanesulfonic acid (MSA), rigid-rod polymer, such as poly(p-phenylenebenzobisoxazole) (PBO), phase separates into undesirable aggregates. The coagulant and the method of its introduction have been found to exert a strong impact on the final film morphology. A quench of a PBO solution in MSA into a water bath results in a three dimensional interconnected network of PBO, while a slower introduction of water results in a more amorphous material. A computer simulation program of the coagulation process has been developed to better understand the different structures emerging from coagulation processing of molecular composites. The simulation results are in qualitative agreement with the experimental observations.