A quantitative and text-based characterization of big data research

Gupta, Vishal; Singh, Vivek Kumar; Ghose, Udayan; Mukhija, Pankaj

doi:10.3233/jifs-179016

Cited by 13 publications

(14 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While DS was more significant in early years for about five decades, BD made a leap during the recent decade. Interestingly, these results are in agreement with a previous study that retrieved results from Scopus for the years 2010-16 (Gupta et al 2019).…”

Section: Methodssupporting

confidence: 93%

“…While WoS covers selected conference proceedings, coverage is partial for this type of document, thereby limiting our analysis mostly to peerreviewed journal papers with minor coverage of conference proceedings. Observing the results of an earlier study, we conclude that the trajectory of conference publications is steeper than that of journal papers (Gupta et al 2019). The implication for our research is to say that our findings are conservative and that including more conference proceedings would result in stronger effects in the same direction.…”

Section: Limitationssupporting

confidence: 41%

“…A more recent study investigated the interdisciplinary collaboration in BD research and discovered that the main contributors to the BD research are Computer Science, Engineering, and Business and Economics, and that research communities on the subject are formed (Hu and Zhang 2017). Another recent study covering seven years of BD publications showed the strong dynamics of journal and even more so, of conference publications in the field, attributing much of the activity to the broad interest in BD from various research fields, as well as to the strong interest of large and powerful countries such as China and the U.S. (Gupta et al 2019). A comment in Nature explained the complexity of BD, the need for new algorithms to aggregate the deluge of data, and the future contribution of BD to various areas of research (Mattmann 2013).…”

Section: Literature Reviewmentioning

confidence: 99%

See 2 more Smart Citations

The evolution of data science and big data research: A bibliometric analysis

Raban

Gordon

2020

Scientometrics

View full text Add to dashboard Cite

In this study the evolution of Big Data (BD) and Data Science (DS) literatures and the relationship between the two are analyzed by bibliometric indicators that help establish the course taken by publications on these research areas before and after forming concepts. We observe a surge in BD publications along a gradual increase in DS publications. Interestingly, a new publications course emerges combining the BD and DS concepts. We evaluate the three literature streams using various bibliometric indicators including research areas and their origin, central journals, the countries producing and funding research and startup organizations, citation dynamics, dispersion and author commitment. We find that BD and DS have differing academic origin and different leading publications. Of the two terms, BD is more salient, possibly catalyzed by the strong acceptance of the pre-coordinated term by the research community, intensive citation activity, and also, we observe, by generous funding from Chinese sources. Overall, DS literature serves as a theory-base for BD publications.

show abstract

Section: Methodssupporting

confidence: 93%

Section: Limitationssupporting

confidence: 41%

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

The evolution of data science and big data research: A bibliometric analysis

Raban

Gordon

2020

Scientometrics

View full text Add to dashboard Cite

show abstract

“…In cases where x and x is the attribute value in the complete and incomplete data , the correlation coefficient r can be calculated according to the formula (13). (11) The root mean square error (RMSE) is a well-known main criterion used to compare the performance of prediction methods by measuring the difference between the estimated value of a given characteristic and the baseline value. In this case, a value closer to 0 results in better imputation [69,70].…”

Section: Performance Evaluationmentioning

confidence: 99%

“…The phenomenon of these values is pervasive in clinical studies involving large amounts of data (big data). Meanwhile, big data are extremely large datasets that pose storage and analysis challenges using conventional management techniques [11].…”

Section: Introductionmentioning

confidence: 99%

Normalization and outlier removal in class center-based firefly algorithm for missing value imputation

2021

View full text Add to dashboard Cite

A missing value is one of the factors that often cause incomplete data in almost all studies, even those that are well-designed and controlled. It can also decrease a study’s statistical power or result in inaccurate estimations and conclusions. Hence, data normalization and missing value handling are considered the major problems in the data pre-processing stage, while classification algorithms are adopted to handle numerical features. In cases where the observed data contained outliers, the missing value estimated results are sometimes unreliable or even differ greatly from the true values. Therefore, this study aims to propose the combination of normalization and outlier removals before imputing missing values on the class center-based firefly algorithm method (ON + C3FA). Moreover, some standard imputation techniques like mean, a random value, regression, as well as multiple imputation, KNN imputation, and decision tree (DT)-based missing value imputation were utilized as a comparison of the proposed method. Experimental results on the sonar dataset showed normalization and outlier removals effect in the methods. According to the proposed method (ON + C3FA), AUC, accuracy, F1-Score, Precision, Recall, and AUC-PR had 0.972, 0.906, 0.906, 0.908, 0.906, 0.61 respectively. The result showed combining normalization and outlier removals in C3-FA (ON + C3FA) was an efficient technique for obtaining actual data in handling missing values, and it also outperformed the previous studies methods with r and RMSE values of 0.935 and 0.02. Meanwhile, the Dks value obtained from this technique was 0.04, which indicated that it could maintain the values or distribution accuracy.

show abstract