Using TF-IDF n-gram and Word Embedding Cluster Ensembles for Author Profiling

This paper presents our approach and results for the 2017 PAN Author Profiling Shared Task. Language-specific corpora were provided for four langauges: Spanish, English, Portuguese, and Arabic. Each corpus consisted of tweets authored by a number of Twitter users labeled with their gender and the specific variant of their language which was used in the documents (e.g. Brazilian or European Portuguese). The task was to develop a system to infer the same attributes for unseen Twitter users. Our system employs an ensemble of two probabilistic classifiers: a Logistic regression classifier trained on TF-IDF transformed n-grams and a Gaussian Process classifier trained on word embedding clusters derived for an additional, external corpus of tweets.

[1]  Brendan T. O'Connor,et al.  Part-of-Speech Tagging for Twitter: Annotation, Features, and Experiments , 2010, ACL.

[2]  David A. Freedman,et al.  Statistical Models: Theory and Practice: References , 2005 .

[3]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[4]  Nikolaos Aletras,et al.  An analysis of the user occupational class through Twitter content , 2015, ACL.

[5]  Petr Sojka,et al.  Software Framework for Topic Modelling with Large Corpora , 2010 .

[6]  Kalina Bontcheva,et al.  User profiling with geo-located posts and demographic data , 2016, NLP+CSS@EMNLP.

[7]  Yoram Bachrach,et al.  Studying User Income through Language, Behaviour and Affect in Social Media , 2015, PloS one.

[8]  J. MacQueen.  Some methods for classification and analysis of multivariate observations , 1967 .

[9]  Jeffrey Dean,et al.  Efficient Estimation of Word Representations in Vector Space , 2013, ICLR.

[10]  Margaret L. Kern,et al.  Personality, Gender, and Age in the Language of Social Media: The Open-Vocabulary Approach , 2013, PloS one.

[11]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[12]  Benno Stein,et al.  Overview of PAN'17 - Author Identification, Author Profiling, and Author Obfuscation , 2017, CLEF.

[13]  Benno Stein,et al.  Improving the Reproducibility of PAN's Shared Tasks: - Plagiarism Detection, Author Identification, and Author Profiling , 2014, CLEF.

[14]  Kalina Bontcheva,et al.  Topic Models and n-gram Language Models for Author Profiling - Notebook for PAN at CLEF 2015 , 2015, CLEF.

[15]  Brendan T. O'Connor,et al.  Demographic Dialectal Variation in Social Media: A Case Study of African-American English , 2016, EMNLP.

[16]  Ingemar J. Cox,et al.  Inferring the Socioeconomic Status of Social Media Users Based on Behaviour and Language , 2016, ECIR.

[17]  Benno Stein,et al.  Overview of the 5th Author Profiling Task at PAN 2017: Gender and Language Variety Identification in Twitter , 2017, CLEF.