A Personalized Defection Detection and Prevention Procedure based on the Self-Organizing Map and Association Rule Mining: Applied to Online Game Site

Customer retention is an increasinglypressing issue in today's competitiveenvironment. This paper proposes a personalizeddefection detection and prevention procedurebased on the observation that potentialdefectors have a tendency to take a couple ofmonths or weeks to gradually change theirbehaviour (i.e., trim-out their usage volume)before their eventual withdrawal. For thispurpose, we suggest a SOM (Self-Organizing Map)based procedure to determine the possiblestates of customer behaviour from pastbehaviour data. Based on this staterepresentation, potential defectors aredetected by comparing their monitoredtrajectories of behaviour states with frequentand confident trajectories of past defectors.Also, the proposed procedure is extended toprevent the defection of potential defectors byrecommending the desirable behaviour state forthe next period so as to lower the likelihoodof defection. For the evaluation of theproposed procedure, a case study has beenconducted for a Korean online game site. Theresult demonstrates that the proposed procedureis effective for defection prevention andefficiently detects potential defectors withoutdeterioration of prediction accuracy whencompared to that of the MLP (Multi-LayerPerceptron) neural networks.     

Homogeneous 2D MoTe2 p–n Junctions and CMOS Inverters formed by Atomic‐Layer‐Deposition‐Induced Doping

Recently, α‐MoTe₂, a 2D transition‐metal dichalcogenide (TMD), has shown outstanding properties, aiming at future electronic devices. Such TMD structures without surface dangling bonds make the 2D α‐MoTe₂ a more favorable candidate than conventional 3D Si on the scale of a few nanometers. The bandgap of thin α‐MoTe₂ appears close to that of Si and is quite smaller than those of other typical TMD semiconductors. Even though there have been a few attempts to control the charge‐carrier polarity of MoTe₂, functional devices such as p–n junction or complementary metal–oxide–semiconductor (CMOS) inverters have not been reported. Here, we demonstrate a 2D CMOS inverter and p–n junction diode in a single α‐MoTe₂ nanosheet by a straightforward selective doping technique. In a single α‐MoTe₂ flake, an initially p‐doped channel is selectively converted to an n‐doped region with high electron mobility of 18 cm² V^?1 s^?1 by atomic‐layer‐deposition‐induced H‐doping. The ultrathin CMOS inverter exhibits a high DC voltage gain of 29, an AC gain of 18 at 1 kHz, and a low static power consumption of a few nanowatts. The results show a great potential of α‐MoTe₂ for future electronic devices based on 2D semiconducting materials.     

Semiautomated Identification and Classification of Customer Complaints

This paper examines the feasibility of extracting useful information from customer comments using a Naïve Bayes classifier. This was done for a database, obtained from a large Korean mobile telephone service provider, of 533 customer calls to call centers in 2009. After eliminating calls not containing customer complaints or comments, the remaining 383 comments were classified by an expert panel into four domains and 27 complaint categories. The four domains were Transaction‐related (189 comments, 49%), Product‐related (120 comments, 31%), Customer Service or Support‐related (38 comments, 10%) and Customer Outreach and Marketing‐related (36 comments, 9%). The comments were then randomly assigned to either a training set (257 cases, 67%) or test set (126 cases, 33%). The training set was used to develop a Naïve Bayes classifier that correctly predicted the domain 75% of the time and the specific subcategory 51% of the time for the test set. Prediction accuracy was strongly related to prediction strength for both sets of predictions, suggesting that simple filtering strategies where difficult to understand comments are flagged for expert review and easy comments are automatically classified are both technically feasible and likely to be practically valuable. Several strong predictors were also identified that corresponded to categories more detailed than those originally assigned. © 2012 Wiley Periodicals, Inc.     

Highly Stretchable and Sensitive Strain Sensors Using Fragmentized Graphene Foam

Stretchable electronics have recently been extensively investigated for the development of highly advanced human‐interactive devices. Here, a highly stretchable and sensitive strain sensor is fabricated based on the composite of fragmentized graphene foam (FGF) and polydimethylsiloxane (PDMS). A graphene foam (GF) is disintegrated into 200–300 μm sized fragments while maintaining its 3D structure by using a vortex mixer, forming a percolation network of the FGFs. The strain sensor shows high sensitivity with a gauge factor of 15 to 29, which is much higher compared to the GF/PDMS strain sensor with a gauge factor of 2.2. It is attributed to the great change in the contact resistance between FGFs over the large contact area, when stretched. In addition to the high sensitivity, the FGF/PDMS strain sensor exhibits high stretchability over 70% and high durability over 10 000 stretching‐releasing cycles. When the sensor is attached to the human body, it functions as a health‐monitoring device by detecting various human motions such as the bending of elbows and fingers in addition to the pulse of radial artery. Finally, by using the FGF, PDMS, and μ‐LEDs, a stretchable touch sensor array is fabricated, thus demonstrating its potential application as an artificial skin.     

Natural Palm Olein Polyol as a Replacement for Polyether Polyols in Viscoelastic Polyurethane Foam

The impact of replacing three polyether polyols with different levels of a single palm olein‐based natural oil polyol (NOP) was systematically correlated with the changes in foaming reactivity, cell structure, physico‐mechanical properties, and morphology of viscoelastic (VE) foams. The data show that replacing the polyether polyols with the NOP slightly increased the rate of the foaming reactivity. Increasing the NOP content resulted in increased cell size and cells remained fully open. Increased NOP content contributed to higher load bearing properties of VE foam, which can be attributed to higher functionality of NOP compared to polyether polyols. Addition of the NOP slightly increased the resilience of the foams, however, the hysteresis which is the measure of energy absorption remained mostly unaffected. Age properties, characterized by dry and humid compression sets, were mostly unaffected by the replacement of the polyether polyol with the NOP. The addition of NOP did not impact the morphology of the VE foam polymer matrix, which appears to retain a low degree of hard and soft segment domain separation. Overall, the results demonstrate a feasibility that the NOP can be used to partially replace the polyether polyols in VE polyurethane foams without significant impact on the functional performance.     

Urethane‐forming reaction kinetics and catalysis of model palm olein polyols: Quantified impact of primary and secondary hydroxyls

Model palm olein natural oil polyols (NOPs) with varying ratios of primary to secondary hydroxyls were synthesized, characterized, and evaluated in reaction kinetics study with isocyanate in formation of polyurethanes. Reaction rate constants and activation energies associated with primary and secondary hydroxyls of NOPs were quantified. The kinetic study in toluene shows that the NOP containing primary hydroxyls have three times higher reaction rate constants in noncatalyzed reaction with 4,4′‐diphenylmethane diisocyanate (4,4′‐MDI) compared to the model NOP containing only secondary hydroxyls, which is associated with higher activation energy of secondary hydroxyls. However, the difference in reaction rate constants of primary and secondary hydroxyls in NOPs diminished in the reactions catalyzed with dibutyltin dilaurate. Bulk polymerization reaction confirms the kinetics results in toluene, showing that the model NOP containing primary hydroxyls reached gel time at a faster rate. Evaluation of elastomers from bulk polymerization shows low degree of phase separation of hard and soft segments for elastomers based on the model NOPs. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2016 , 133, 42955.     

Wetting characteristics of a water droplet on solid surfaces with various pillar surface fractions under different conditions

A numerical study was carried out using a molecular dynamics program to examine the wetting characteristics of nano-sized water droplets on surfaces with various pillar surface fractions under different conditions. Square-shaped pillars had surface fractions that increased from 11.1 % to 69.4 %. The pillars had 4 different heights and 3 different surface energies. When the pillar surface fraction changed, the contact angle of a water droplet also changed due to the attraction between the droplet and the pillar surface or the inner attraction of the water molecules. The pillar height also has different effects on the water droplet depending on the magnitude of surface energy.     

Preparation of highly conductive reduced graphite oxide/poly(styrene-co-butyl acrylate) composites via miniemulsion polymerization

Polymer/reduced graphite oxide (rGO) composite nanoparticles with a high electrical conductivity were synthesized using the miniemulsion polymerization technique. The rGO was modified with a reactive surfactant, 2-acrylamido-2-methyl-1-propanesulfonic acid (AMPS), to facilitate monomer intercalation into the rGO nanogalleries. The AMPS-modified rGO was emulsified in the presence of styrene (St) and butyl acrylate (BA) monomers, and the stable miniemulsion was polymerized to form poly(St-co-BA)/rGO composite latex nanoparticles. The transition in the composite nanoparticles from an electrical insulator to an electrical conductor occurred at an rGO content of 10 wt% (relative to the monolayer content), yielding an electrical conductivity of 0.49 S/cm. The electrical conductivity of the composite nanoparticles reached 2.22 S/cm at 20 wt% rGO, yielding a much better conductivity than other polymer composites prepared using a GO filler. Importantly, the miniemulsion polymerization method for fabricating poly(St-co-BA)/rGO composite nanoparticles is easy, green, low-cost, and scalable, providing a universal route to the rational design and engineering of highly conductive polymer composites.     

Gallic Acid Grafted Chitosan Has Enhanced Oxidative Stability in Bulk Oils

Gallic acid (GA) was grafted in chitosan and the effects of GA grafted chitosan (GA‐g‐CS) on the oxidative stability in bulk oil was tested at 60 and 140 °C. To text oxidative stability in oils, headspace oxygen content, conjugated dienoic acid (CDA) value, p‐anisidine value (p‐AV), and acid value were determined. Chitosan itself did not show antioxidative or prooxidative effects in oils at 60 °C. However, GA‐g‐CS and GA acted as antioxidants at 60 °C. At 140 °C heating with moisture supplied condition, different results were observed. GA‐g‐CS acted as antioxidants based on the results of CDA and p‐AV. However, chitosan showed the highest oxidative stability based on results of acid value and brown color formation at 140 °C. This could be due to reduction of moisture content by chitosan. GA was continuously released from GA‐g‐CS in bulk oil. This might have provided extra antioxidant activities to oils.