Research Picks Weekly – 8 Feb 2018

Sweet solution to carbon spheres

Carbon particles of regular spherical shape are useful in a wide range of technologies. An aqueous heat treatment, hydrothermal, process can be used to convert natural glucose into such particles with diameters in the range 90 to 490 nanometres. The process takes place at between 170 and 190 degrees Celsius over the course of 1 to 6 hours. The resulting carbon spheres can then be hardened by carbonizing them. The size of the hard carbon spheres produced increases exponentially with reaction temperature and time, but in almost linear fashion in terms of glucose concentration. “The results provide a route to controllably synthesise the carbon spheres in an array of sizes and dispersity by adjusting parameters,” the researchers report.

“Size controllable synthesis of hard carbon spheres from aqueous D-glucose”, Guanggui Cheng; Joseph Cremaldi; Jianning Ding; Yang Su; Yueheng Zhang; Noshir S. Pesika; Ying Wang, Int J Mater Struct Integ, 2017, 11, 213 – 228; http://dx.doi.org/10.1504/IJMSI.2017.10010801


Crowdflower data blooms

It should be possible to data mine updates on social media sites, such as Twitter, to check the veracity of information on another online source, such as Wikipedia, according to new research from a team in Italy. The researchers propose a model using crowdsourcing to disambiguate and decide on the accuracy of a given Wikipedia page based on Twitter updates associated with specific terms. The team used Crowdflower instead of Amazon Mechanical Turk for its flexibility and the greater number of channels. By then manually removing “spam” the team was able to home in with a confidence level of 1 on each information spot.

“A crowdsourced system for user studies in information extraction”, Zohreh Khojasteh-Ghamari, Int J Knowledge Eng Soft Data Paradigms, 2017, 6; https://doi.org/10.1504/IJKESDP.2017.089506


Catch a train

A new algorithm can more accurately predict when the next train will arrive at the station by carrying out a statistical analysis of that day’s arrivals and departures and those for trains earlier in the week. The improvement amounts to a better prediction error reduced from 12 seconds to 3 seconds. Such a signal boost may not seem important given the timescales on which passengers board and alight trains, however, improved timings could be used to significantly improve the efficiency of regenerative braking systems employed by modern trains so that braking is optimised according to the trains ahead of each other. The kinetic energy of a braking train behind a departing train can be used to boost the power available to the departing train and thus reduce energy wasted in the braking process. The algorithm employs “portfolio theory” to work its timetable magic.

“Portfolio theory application to prediction correction of train arrival times” Takaaki Yamada and Tatsuhiro Sato, Int J Computat Intelligence Studies, 2017, 6; https://doi.org/10.1504/IJCISTUDIES.2017.089519


Take a bung

Bribery and corruption are on the increase in the developing world where it is estimated that the equivalent of US$400 billion is paid illicitly and illegally to the political elite in those nations. The scale of such corruption suggests that multinational, “Western” companies must be complicit in such transactions in the sense that they essentially facilitate the laundering and banking of this money and ultimately how it is spent. Researchers from Nigeria and the UK suggest that multinational enterprise culture and accounting practices mean that the drive for higher profits is at almost any cost and as such is not constrained by the rule of law nor regulatory activities. The team’s evidence suggests that not only are the multinationals complicit they are actively gaining a competitive advantage by being directly engaged in bribery and corruption. The team makes various suggestions for reform.

“The culpability of accounting practice in promoting bribery and corruption in developing countries” Olatunde Julius Otusanya, Sarah Lauwo Amal Hayati Ahmad-Khair, Int J Econ Account, 2018; https://doi.org/10.1504/IJEA.2017.089387