Exploring latent browsing graph for question answering recommendation

In this paper, we develop a framework of Question Answering Pages (referred to as QA pages) recommendation. Our proposed framework consists of the two modules: the off-line module to determine the importance of QA pages and the on-line module for on-line QA page recommendation. In the off-line module, we claim that the importance of QA pages could be discovered from user click streams. If the QA pages are of higher importance, many users will click and spend their time on these QA pages. Moreover, the relevant relationships among QA pages are captured by the browsing behavior on these QA pages. As such, we exploit user click streams to model the browsing behavior among QA pages as QA browsing graph structures. The importance of QA pages is derived from our proposed QA browsing graph structures. However, we observe that the QA browsing graph is sparse and that most of the QA pages do not link to other QA pages. This is referred to as a sparsity problem. To overcome this problem, we utilize the latent browsing relations among QA pages to build a QA Latent Browsing Graph. In light of QA latent browsing graph, the importance score of QA pages (referred to as Latent Browsing Rank) and the relevance score of QA pages (referred to as Latent Browsing Recommendation Rank) are proposed. These scores demonstrate the use of a QA latent browsing graph not only to determine the importance of QA pages but also to recommend QA pages. We conducted extensive empirical experiments on Yahoo! Asia Knowledge Plus to evaluate our proposed framework.     

An empirical study of design-of-experiment data mining for yield-loss diagnosis for semiconductor manufacturing

To maintain competitive advantages, semiconductor industry has strived for continuous technology migrations and quick response to yield excursion. As wafer fabrication has been increasingly complicated in nano technologies, many factors including recipe, process, tool, and chamber with the multicollinearity affect the yield that are hard to detect and interpret. Although design of experiment (DOE) is a cost effective approach to consider multiple factors simultaneously, it is difficult to follow the design to conduct experiments in real settings. Alternatively, data mining has been widely applied to extract potential useful patterns for manufacturing intelligence. However, because hundreds of factors must be considered simultaneously to accurately characterize the yield performance of newly released technology and tools for diagnosis, data mining requires tremendous time for analysis and often generates too many patterns that are hard to be interpreted by domain experts. To address the needs in real settings, this study aims to develop a retrospective DOE data mining that matches potential designs with a huge amount of data automatically collected in semiconductor manufacturing to enable effective and meaningful knowledge extraction from the data. DOE can detect high-order interactions and show how interconnected factors respond to a wide range of values. To validate the proposed approach, an empirical study was conducted in a semiconductor manufacturing company in Taiwan and the results demonstrated its practical viability.     

Optimization of membrane gas separation systems using genetic algorithm

Genetic algorithm is applied for the optimization of the membrane gas separation systems. Air separation for enriched oxygen production is the selected system for investigation. Optimizations for single and triple objective functions are studied. The optimization problem involves the selection of the optimal system configurations from three alternatives, including continuous membrane column (CMC), single stripper permeator (SSP), and two stripper in series permeator (TSSP), as well as the optimal operating conditions. Models of the three configurations and the genetic algorithm procedure are computerized. The objective functions discussed are the Rony separation index, power consumption per unit equivalent pure oxygen, and the membrane area. Both high-pressure and low-pressure (vacuum) operation modes are optimized and the effects of different oxygen product purity and feed rate are analyzed. For single objective function optimization, the solutions obtained using genetic algorithm are slightly inferior in one case but superior in other cases compared to those by pure mathematical optimization methods. For triple objective function optimization, the Pareto plots presenting multiple trade-off solutions are generated. In general, compared to high-pressure operation mode, the product recovery and power consumption for low-pressure operation mode are lower. For almost all the cases studied, CMC configuration with its high flexibility appears in the optimal solutions.     

Modified Indigo Method For Gaseous And Aqueous Ozone Analyses

The indigo method developed by Bader and Hoigné for aqueous ozone analysis was modified to allow for both gaseous and aqueous ozone determination. Gas or water samples were extracted with a gas-tight syringe containing a known volume of indigo reagent. The modified procedure provided a more consistent basis for gaseous and aqueous ozone determination allowing for more accurate ozone mass balance calculations. Direct gaseous ozone UV absorbance with molar absorptivity of 3,000 M^?1cm^?1 at 258 nm was used as primary standard to determine the molar absorptivity of the indigo reagent. The molar absorptivity of indigo reagent, assuming a 1:1 stoichiometric ratio for the reaction between indigo and ozone, was determined to be 23,150 ± 80 M^?1cm^?1, or approximately 16 percent higher than that of 20,000 M^?1cm^?1 suggested by Bader and Hoigné. An independently calibrated membrane-electrode ozone monitor showed good correlation with indigo method results using the molar absorptivity value determined in this study. The apparent molar absorptivity of aqueous ozone at the wavelength of 258 nm measured by the modified indigo method increased from 2,400 to 3,600 M^?1cm^?1 in the investigated ozone concentration range of 0.4 to 11.0 mg/L. This variation might have been caused by the inherent interference of unidentified ozone byproducts, which presence was supported with scanning spectra in the wavelength range of 200 to 300 nm.     

Gas and vapor sorption in polymers just below Tg

Generally, sorption isotherms for gases like CO₂ in glassy polymers are concave to the pressure axis, whereas in the rubbery state these isotherms are linear for gases or sometimes convex to the pressure axis for more condensable vapors. Examples of CO₂ isotherms are reported here that show at low pressure the curvature characteristic of glasses and then become linear at higher pressures. This is observed when the glass transition temperature T_g is not much greater than the observation temperature T, and plasticization of the polymer by sorbed CO₂ causes T_g to become equal to T within the range of pressures employed in the isotherm measurement. For the sorption of vapors in glassy polymers, this can lead to sigmoidal isotherms, as discussed using an illustration from the literature.     

Properties and pervaporation separation of hydroxyl‐terminated polybutadiene‐based polyurethane/poly(methyl metharcylate) interpenetrating networks membranes

Hydroxyl‐terminated polybutadiene (HTPB), 4,4′‐dicyclohexyl methane diiscyanate (H₁₂MDI), and 1,4‐butane diol are used to synthesize polyurethane (PU) solutions by two‐stage process. Interpenetrating networks (IPNs) of HTPB‐based PU and poly(methyl methacrylate) (PMMA) with HTPB/MMA (wt/wt % ratio) = 2.0, 1.5, 1.0, 1.5, 0.8, and 0.6, which are designated as IPN1 to IPN5, respectively, are synthesized by sequential polymerization technique. Thermal properties, tensile strength, and contact angle of membranes increase with the increase of MMA content, while the elongation of membranes show the reverse trend. Characterization of membranes are investigated by C?C/C?O absorption ratio and infrared absorption frequency shiftment. These PU and IPN membranes are used for the separation of ethanol/water and isopropanol/water solution by pervaporation test. IPN3 membrane possesses the largest pervaporation permeability and the separation factor. The pervaporation results of ethanol/water feed has the same trend as that of isopropyl alcohol (IPA)/water solution. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 2007     

Numerical simulation for Fuzzy-PID controllers and helping EP reproduction with PSO hybrid algorithm

Many Fuzzy-PID controller schemes used in industry today are based on some sort of simplified fuzzy reasoning methods and PID parameters. We present a design for Fuzzy-PID controllers using a novel PSO-EP-based hybrid algorithm. We succeed in making mathematical calculations and in encouraging EP reproduction with PSO. The main advantage of our design is that the integration of the PSO-EP-based hybrid algorithm structure generates new parameters for the Fuzzy-PID control schemes. The proposed algorithm is an improved variant of PSO, a relatively recently introduced stochastic optimization strategy for Fuzzy-PID controllers that is investigated in this study. The function of Fuzzy-PID controllers is illustrated by means of a model of the induction motor control system and a higher-order numerical model.