Examining the performance of classification algorithms for imbalanced data sets in web author identification

Vorobeva, Alisa

doi:10.1109/fruct-ispit.2016.7561554

Cited by 12 publications

(3 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More studies on author attribution on imbalanced data proposed that uses many short text samples for the minority classes and less and longer text samples for the majority classes [36]. Therefore, it is necessary to take into account the size of the samples used for testing to evaluate the accuracy when performing the identification process and most studies in this area proved that there is no mechanism and solid to provide the appropriate size of the samples for the testing to be applied in biometrics during the identification process.…”

Section: Stylometric Features In Long Textmentioning

confidence: 99%

Surveying the Development of Authorship Identification of Text Messages

Altamimi¹,

Alotaibi²,

Alruban³

2019

IJICR

View full text Add to dashboard Cite

People typically use multiple messaging systems and send text messages concurrently by different messaging systems such as SMS, email, Twitter, and Facebook. This means that these forms of messaging systems are highly integrated into people's everyday activities. Moreover, forensic science has had many challenges in many different kinds of crimes, differing from physical crimes to computer mediated crime activities. Identification of suspects has become crucial for law enforcement due in particularly to the anonymity that the internet and associated services provide and identify the ownership of messages. In this paper, survey the development of existing author identification have been done by establishing a comprehensive review of author identification to determining the core approaches and techniques utilized. Furthermore, this study also examines possible challenges in author identification and points to some of the open problems which need to be tackled.

show abstract

Section: Stylometric Features In Long Textmentioning

confidence: 99%

Surveying the Development of Authorship Identification of Text Messages

Altamimi¹,

Alotaibi²,

Alruban³

2019

IJICR

View full text Add to dashboard Cite

show abstract

“…Before we begin our investigation of algorithm performance we need to understand the basics of what is classification and regression and why are they used. Classification in machine learning essentially means that to place a new observation in a set of categories already predefined based on the training dataset [3]. So in classification we actually group the output variables into different corresponding classes.…”

Section: Introductionmentioning

confidence: 99%

Measuring Performance of Selected Algorithms used for Classification and Regression when Applied Against a Standard Dataset

Bharadwaj¹,

Sunitha²,

Shashidhara³

2017

IJERT

View full text Add to dashboard Cite

Data mining algorithms are often used to extract useful information from datasets, which give us deeper insights into what we are focussing on to get from the data. This paper aims at measuring the performance of the few selected algorithms namely, Bayesian Generalized Linear Model, Generalized Linear Model, k-Nearest Neighbours and Partial Least Squares. The various performance parameters measured include sensitivity, specificity, root mean squared error (RMSE), R Squared etc. The dataset used would be from the machine learning library present in R studio namely, mlbench library and the dataset being Pima Indians Diabetes.

show abstract

“…The numerous real-world applications are affected by class imbalance problem wherein the number of samples in one class is very marginal compared to other classes [8][9][10][11]. Issues in fields related to software defect detection [12], threat supervision, medical judgment, web author identification [13] and similar have drawn attention towards concerns of multi-class imbalanced data sets. The representation of boundaries in imbalanced data sets is a difficult concern for learning algorithms.…”

mentioning

confidence: 99%