Statistics-Based Outlier Detection and Correction Method for Amazon Customer Reviews

People nowadays use the internet to project their assessments, impressions, ideas, and observations about various subjects or products on numerous social networking sites. These sites serve as a great source to gather data for data analytics, sentiment analysis, natural language processing, etc. Conventionally, the true sentiment of a customer review matches its corresponding star rating. There are exceptions when the star rating of a review is opposite to its true nature. These are labeled as the outliers in a dataset in this work. The state-of-the-art methods for anomaly detection involve manual searching, predefined rules, or traditional machine learning techniques to detect such instances. This paper conducts a sentiment analysis and outlier detection case study for Amazon customer reviews, and it proposes a statistics-based outlier detection and correction method (SODCM), which helps identify such reviews and rectify their star ratings to enhance the performance of a sentiment analysis algorithm without any data loss. This paper focuses on performing SODCM in datasets containing customer reviews of various products, which are (a) scraped from Amazon.com and (b) publicly available. The paper also studies the dataset and concludes the effect of SODCM on the performance of a sentiment analysis algorithm. The results exhibit that SODCM achieves higher accuracy and recall percentage than other state-of-the-art anomaly detection algorithms.     

Observation of carbon‐facilitated phase transformation of titanium dioxide forming mixed‐phase titania by confocal Raman microscopy

Confocal Raman microscopy, a relatively new and advanced technique, is found to be suitable for imaging the chemical morphology below the submicrometer scale. It has been employed to probe the phase transformation of carbon‐containing titania (TiO₂) nanopowder and titania thin film subjected to laser annealing. The observation of phase transformation from the anatase phase to the rutile phase at high laser power annealing is attributed to carbon inclusion inside or on the surface of titania. Upon annealing, carbon could react with the oxygen of titania and create oxygen vacancies favoring the transformation from the anatase to the rutile phase. This study provides evidence for the carbon‐assisted phase transformation for creating carbon‐containing mixed‐phase titanium dioxide by laser annealing. We explicitly focus on the presence of carbon in the phase transformation of TiO₂ using confocal Raman microscopy. In all of the investigated samples, mixed anatase/rutile phases with carbon specifically was found at the rutile site. X‐ray diffraction (XRD), scanning electron microscopy (SEM) and energy‐dispersive spectroscopy (EDS) studies have been performed in addition to Raman mapping to verify the mixed‐phase titania formation. Copyright © 2010 John Wiley & Sons, Ltd.     

Fluctuation study of pionisation in ultrarelativistic nucleus-nucleus interaction

The scaled factorial moments and the multifractal moments have been investigated in differentη-intervals to study the dynamical fluctuation of pions produced in 200 AGeV³²S-Ag/Br interaction. In order to investigate the detail characteristics of intermittency behaviour, theF-moments are extracted up to the eighth order of moments in differentM-intervals. The analysis indicates a non-thermal phase transition and different regime of particle production during the hadronisation process.     

Effects of final-state interaction and screening on strange- and heavy-quark production

Final-state interaction and screening have a great influence on $q\bar q$ production cross sections, which are important quantities in many problems in quark-gluon plasma physics. They lead to an enhancement of the cross section for a $q\bar q$ color-singlet state and a suppression for a color-octet state. The effects are large near the production threshold. The presence of screening gives rise to resonances for $q\bar q$ production just above the threshold at specific plasma temperatures. These resonances, especially $c\bar c$ and $b\bar b$ resonances, may be utilized to search for the quark-gluon plasma by studying the temperature dependence of heavy-quark pair production just above the threshold.     

Influence of electrolytic hydrogen on the etch pit density of iron (110) surfaces

Etch pit densities on iron (110) surfaces in sulphuric acid grow linearly with the interfacial hydrogen activity in excess of a critical activity. The hydrogen activity is approximately proportional to the square root of the cathodic current density. At constant cathodic current density the etch pit density increases with temperature and decreases with external stress. Dislocations at which the excess etch pits form penetrate into the iron at a rate proportional to the hydrogen activity and the square root of time. Effects of prior hydrogen deposition on the shape of etch pits are seen at depths greater than the penetration depth of hydrogen generated dislocations. Changes of etch pit shape similar to those produced by hydrogen are also found when external stress is applied.The results are compared to Prussin's theory in which the assumption is made that stresses accompanying diffusion of an impurity are fully relieved by plastic deformation and formation of dislocations for stresses exceeding a critical stress. While some of the predictions of the theory are met by the experiments, the dislocations penetrate into the iron much slower than diffusion of hydrogen, since dislocations cannot move fast enough, i.e. stresses are not fully relieved.     

Reaction48Ca (p,n)48Sc fromE p =1.885 to 5.1 MeV

The total (p, n) reaction cross section for⁴⁸Ca has been measured as a function of proton energy in the energy range 1.885 to 5.100 MeV with an overall resolution of ∼ 2 keV and in ∼ 5 keV energy steps. The fluctutions in fine resolution data have been analysed to determine the average coherence width 〈Γ〉. The excitation function averaged over large energy intervals has been analyzed in terms of the optical model. The isobaric analogue resonances atE _p ∼ 1.95 and 4 MeV have been shape-analyzed to extract the proton partial width and the spectroscopic factorS _n . A comparison of the gross structures observed in ∼ 55 keV averaged excitation function with the predictions of Izumo’s partial equilibrium model has also been made.     

A high spin isomer in 153Eu

A search for new isomers of nanosecond lifetimes were carried out in ¹⁵³Eu via the ¹⁵⁰Nd(⁷Li,xnγ) reaction. The single particle angular momentum alignment and dynamical moment of inertia estimated from the experimental data indicate a configuration change at rotational energy ħω∼ 0.30 MeV. A decrease of B(E2) values is also observed at the same frequency. An isomeric level is identified at an excitation energy of 3100 keV (J^π=35/2⁻) which corresponds to this frequency. The lifetime of the level is found to be 8.6 ± 1.3 nanosecond. Received: 22 April 1997 / Revised version: 26 September 1997     

Response of regional brain glutamate transaminases of rat to aluminum in protein malnutrition

Background  

The mechanism of aluminum-induced neurotoxicity is not clear. The involvement of glutamate in the aluminium-induced neurocomplications has been suggested. Brain glutamate levels also found to be altered in protein malnutrition. Alterations in glutamate levels as well as glutamate-α-decarboxylase in different regions of rat brain has been reported in response to aluminum exposure. Thus the study of glutamate metabolising enzymes in different brain regions of rats maintained on either normal or restricted protein diet may be of importance for understanding the neurotoxicity properties of aluminium.