Machine Learning Models of Text Categorization by Author Gender Using Topic-independent Features

Sboev, Alexander; Litvinova, Tatiana; Gudovskikh, Dmitry; Rybka, Roman; Moloshnikov, Ivan

doi:10.1016/j.procs.2016.11.017

Cited by 50 publications

(32 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The topology of network is taken from the work [15] to determine the sex of the author, it based only on the morphological characteristics of words. A complicated neural network combining CNN, MLP and LSTM includes: 1st, 3rd, 5th CNN layers: Number of convolution kernels to use = 128, the extension of each filter = 2, activation function is ReLU; 2nd, 4th, 6th layers: MaxPooling (pool length = 2); 7th layer: Long-Short Term Memory (output dimension = 128); 8th layer: dropout layer.…”

Section: Pre-trained Modelsmentioning

confidence: 99%

Data-driven Approaches to Author’s Profiling Identification for Russian Texts on Base of Complex Machine Learning Models in Combinations with Siamese Networks

Sboev¹,

Moloshnikov²,

Gudovskikh³

et al. 2018

dtcse

Self Cite

View full text Add to dashboard Cite

Abstract. In this work data-driven approaches to author's profiling identification for Russian texts are investigated on base of a united data corpus. This corpus has been specially collected by crowdsourcing, and currently contains texts from 1161 men and 2043 women. The adaptation of complicated models, based on convolutional neural networks, gradient boosting methods, LSTM, Siamese networks along with different input data and features (morphological data, vector of character n-grams frequencies, Linguistic Inquiry and Word Count and others) to form the vector of derived features in order to identify gender and age of the author of text is described. The method to improve the accuracy using coding by the Siamese network is presented and analyzed.

show abstract

Section: Pre-trained Modelsmentioning

confidence: 99%

Data-driven Approaches to Author’s Profiling Identification for Russian Texts on Base of Complex Machine Learning Models in Combinations with Siamese Networks

Sboev¹,

Moloshnikov²,

Gudovskikh³

et al. 2018

dtcse

Self Cite

View full text Add to dashboard Cite

show abstract

“…The detection of gender, age, and other clues of the author's personality and language has also attracted a great deal of attention (Argamon et al, 2007a;Cheng et al, 2011;Schler et al, 2006;Argamon et al, 2007b;Peersman et al, 2011;Rangel and Rosso, 2013;Simaki et al, 2015aSimaki et al, ,b, 2016Simaki et al, , 2017Sboev et al, 2016;Lins and Gonçalves, 2004). These studies investigate one or more sociodemographic factors, and many of them use data from social media.…”

Section: Related Workmentioning

confidence: 99%

Identifying the Authors’ National Variety of English in Social Media Texts

Simaki¹,

Simakis²,

Paradis³

et al. 2017

RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning

View full text Add to dashboard Cite

In this paper, we present a study for the identification of authors' national variety of English in texts from social media. In data from Facebook and Twitter, information about the author's social profile is annotated, and the national English variety (US, UK, AUS, CAN, NNS) that each author uses is attributed. We tested four feature types: formal linguistic features, POS features, lexicon-based features related to the different varieties, and databased features from each English variety. We used various machine learning algorithms for the classification experiments, and we implemented a feature selection process. The classification accuracy achieved, when the 31 highest ranked features were used, was up to 77.32%. The experimental results are evaluated, and the efficacy of the ranked features discussed.

show abstract

“…In this paper we investigate the task of gender identification of Russian text author in texts with gender deception (i. e. when the authors tried to mimic other gender) in comparison with the same task for texts without deception. While the latter case is widely represented in scientific literature, see for example [1,2,3,4,5,6,7,8], sources for the former case are absent. The closest formulation is issued in [9], where the task of presence of any deception in men and women texts is investigated.…”

Section: Introductionmentioning

confidence: 99%

A comparison of Data Driven models of solving the task of gender identification of author in Russian language texts for cases without and with the gender deception

Sboev

Moloshnikov

Gudovskikh

et al. 2017

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Abstract. In this work we compare several data-driven approaches to the task of author's gender identification for texts with or without gender imitation. The data corpus has been specially gathered with crowdsourcing for this task. The best models are convolutional neural network with input of morphological data (f1-measure: 88%±3) for texts without imitation, and gradient boosting model with vector of character n-grams frequencies as input data (f1-measure: 64% ± 3) for texts with gender imitation. The method to filter the crowdsourced corpus using limited reference sample of texts to increase the accuracy of result is discussed.

show abstract

Machine Learning Models of Text Categorization by Author Gender Using Topic-independent Features

Cited by 50 publications

References 6 publications

Data-driven Approaches to Author’s Profiling Identification for Russian Texts on Base of Complex Machine Learning Models in Combinations with Siamese Networks

Data-driven Approaches to Author’s Profiling Identification for Russian Texts on Base of Complex Machine Learning Models in Combinations with Siamese Networks

Identifying the Authors’ National Variety of English in Social Media Texts

A comparison of Data Driven models of solving the task of gender identification of author in Russian language texts for cases without and with the gender deception

Contact Info

Product

Resources

About