Integrating Visual Analytics Support for Grounded Theory Practice in Qualitative Text Analysis

We present an argument for using visual analytics to aid Grounded Theory methodologies in qualitative data analysis. Grounded theory methods involve the inductive analysis of data to generate novel insights and theoretical constructs. Making sense of unstructured text data is uniquely suited for visual analytics. Using natural language processing techniques such as parts‐of‐speech tagging, retrieving information content, and topic modeling, different parts of the data can be structured and semantically associated, and interactively explored, thereby providing conceptual depth to the guided discovery process. We review grounded theory methods and identify processes that can be enhanced through visual analytic techniques. Next, we develop an interface for qualitative text analysis, and evaluate our design with qualitative research practitioners who analyze texts with and without visual analytics support. The results of our study suggest how visual analytics can be incorporated into qualitative data analysis tools, and the analytic and interpretive benefits that can result.     

A fast algorithm for learning a ranking function from large-scale data sets

We consider the problem of learning the ranking function that maximizes a generalization of the Wilcoxon-Mann-Whitney statistic on the training data. Relying on an $\epsilon$-accurate approximation for the error-function, we reduce the computational complexity of each iteration of a conjugate gradient algorithm for learning ranking functions from O(m2) to O(m2), where m is the number of training samples. Experiments on public benchmarks for ordinal regression and collaborative filtering indicate that the proposed algorithm is as accurate as the best available methods in terms of ranking accuracy, when the algorithms are trained on the same data. However, since it is several orders of magnitude faster than the current state-of-the-art approaches, it is able to leverage much larger training datasets.     

Studies on poly(vinylchloride)-based polymer blends intended for medical applications. Part II: Mechanical properties

Binary polyblends of plasticized poly(vinylchloride) (PPVC1) with acrylonitrile-butadiene rubber (NBR1), compounded graft polymer of vinyl chloride and ethylene-vinyl acetate copolymer (EVAPVC) and ternary blends of PPVC1, NBR1, and EVAPVC were formulated to study mechanical properties. For the “pure” components, elongation at break was found to be in the order PPVC1 < EVAPVC < NBR1. Addition of both EVAPVC and NBR1 resulted in improvement of ultimate elongation and tensile energy to break. PPVC1-NBR1 binary blends exhibited synergistic behavior for both ultimate tensile stress and elongation at break, indicating the presence of appreciable specific interactions between the polymers. About 30% replacement of PPVC1 by NBR1 or EVAPVC resulted in marginal fall in modulus and significant improvement in elongation at break. EVAPVC or blends containing EVAPVC had a tendency to creep. A simple generalized equation suggested for various mechanical properties works satisfactorily. The results obtained have practical implications and indicate that PVC-based polyblends can be investigated further for potential applications in medicine.     

Structural relaxation in poly(ethylene terephthalate) studied by positron annihilation lifetime spectroscopy

Positron lifetime technique was used to study the physical ageing process in poly(ethylene terephthalate) (PET). Positron lifetime results show that the structural relaxation processes in PET encompass two different time regimes, one short and the other long. The relaxation function constructed from the measured o‐Ps intensity I₃ exhibits non‐exponential character, which can be best fitted with two additive exponentials. The Narayanaswamy model (Kohlraush‐William Watt (KWW) function) is invoked to extract the stretching parameter β indicating the extent of deviation from exponential relaxation. Based on the relaxation times, the activation energies calculated seem to label the different kinetic units of PET structure participating in the relaxation process. © 2002 Society of Chemical Industry     

Multi-objective search group algorithm for thermo-economic optimization of flat-plate solar collector

Neural Computing and Applications - This study aims to develop a multi-objective version of the search group algorithm (SGA) called the multi-objective search group algorithm (MOSGA) to help...     

Process-induced residual stresses in compression molded UHMWPE

Non-isothermal cooling during processing causes the development of residual stresses, which are analyzed for compression molded UHMWPE, and affects the dimensional stability. The development of thermal residual stresses was predicted using an incremental stress analysis that included temperature-dependent material properties. Strain gauges were used to measure the residual stresses as layers were removed from a molded disk using a Process Simulated Laminate (PSL) approach. The PSL technique has not previously been applied to a compression molded neat polymer. For initial surface cooling rates of ～ 11°C/min, the model predicted a compressive stress at the bottom surface of 14 MPa and a tensile stress near the center of 2.5 MPa and matched the experimental distribution well. Because the compressive residual stress was 70% of the yield strength (～20 MPa), a lower cooling rate was also tested (2.6°C/min). The maximum tensile and compressive stresses for this cooling rate were, 0.91 MPa and 2.5 MPa, respectively. The model demonstrated its use for predicting thermal residual stresses in compression molded parts, instead of trial-and-error experimentation. UHMWPE is shown to develop residual stresses continually from ～ 120°C to 23°C.     

A comparative study on the influence of alkyl thiols on the structural transformations in P3HT/PCBM and P3OT/PCBM blends

Herein, we present a comparative study on the structural transformations in P3HT/PCBM and P3OT/PCBM blends using octane (C₈) and dodecane (C₁₂) thiols as solvent additives. Addition of dodecane thiol to the blends promotes needle-like PCBM crystal formation. These crystals are found abundant in P3OT/PCBM and less in P3HT/PCBM blend and are identified by SAXS, DSC and SEM measurements. Using polymer of two different tail lengths, the SAXS results provide the first experimental support on the recent findings on fullerene intercalation around the crystalline boundaries in a semi-crystalline polymer. The UV–Visible measurements reveal that the absorption maximum show an appreciable red-shift and photoluminescence spectra infers that the donor/acceptor interface is mostly retained, upon dodecane thiol addition. These needle-like PCBM crystals obtained by annealing free approach in the blends and their dependence on the alkyl tail length of the polymer, would pave way for more effective design of organic photovoltaic devices.     

Isothermal Crystallization Kinetics of Poly (Ethylene Terephthalate) – Cellulose Acetate Blends

Summary The isothermal crystallization exotherm was determined at three temperatures for two blends of bottle regrind Poly(ethylene terephthalate) (PET) with Cellulose Acetate (CA) as derived by an extrusion process, as well as for the pristine PET. It was observed that the addition of 10 weight-percent CA increased the crystallization rate of PET manifold, which resulted in improved moldability with faster cycle times. The crystallization was observed to be very fast with initiation times of the order of a few seconds, perhaps due to the impurities present in the bottle regrind PET and the synergistic effects of the CA particles. The fast rate of crystallization and high melting point of PET (250 C), and the degradation of CA at temperatures higher than 260 C made extrusion of PET – CA blends difficult. However, increasing the CA content in the blend to 30% decreased the rate of crystallization of PET resulting in an improved extrusion process, but slower cycle times.