Computational Models of Human Personality for Natural Language Processing Applications
Overview
The project focuses on the computational treatment of human personality from both natural language understanding and generation perspectives. Based on the well-established Big Five model of personality traits, the project proposes the collection and annotation of a basic linguistic-computational resource - which can be seen as a parallel corpus of texts and personality inventories - in order to map the relationships between personality traits and a broad range of linguistic phenomena, and then use this resource to build both computational models of personality recognition from text, and models of text generation based on target personality profiles.
Current status
The project ran from March 2017 to February 2019 under FAPESP grant nr. 2016/14223-0.
Deliverables
The b5 corpus of controlled and Facebook texts accompanied by personality inventories is available for download here, and it may be freely reused for research purposes. The corpus is described in the LREC-2018 paper by Ramos et. al. - see below for details. The corpus is available for download here,
The FeedbackRE corpus is an extension of both GRE3D3 (Dale & Viethen, 2009) and the b5-ref subcorpus, and it contains descriptions elicited under conditions with and without automated and human feedback alike.
Selected publications
Deutsch, Caio ; Ivandré Paraboni (2023) Authorship Attribution using Author Profiling Classifiers. Natural Language Engineering 29(1), pp. 110-137. 10.1017/S1351324921000383.
Custódio, José Eleandro ; Ivandré Paraboni (2021) Stacked Authorship Attribution of Digital Texts. Expert Systems with Applications, vol. 176. 10.1016/j.eswa.2021.114866.
Rocha, Danillo da Silva; Ivandré Paraboni (2020) Building Referring Expression Corpora with and without feedback. Language Resources and Evaluation 54(4), pp. 875-891. 10.1007/s10579-020-09497-2.
Stavracas Neto, Georges Basile ; Ivandré Paraboni (2020) Reescrita sentencial baseada em traços de personalidade. Linguamática 12(1), pp. 49-61.
Ramos, Ricelli M. S. ; Danielle S. Monteiro; Ivandré Paraboni (2020) Personality-dependent content selection in natural language generation systems. Journal of the Brazilian Computer Society 26:2. 10.1186/s13173-020-00096-1
Dias, Rafael Felipe Sandroni ; Ivandré Paraboni (2020) Cross-domain Author Gender Classification in Brazilian Portuguese.
12th Language Resources and Evaluation Conference (LREC-2020). pp. 1220-1227, Marseille, France.
Santos, W. R. ; Ricelli M. S. Ramos ; Ivandré Paraboni (2020) Computational Personality Recognition from Facebook text:psycholinguistic features, words and facets. New Review of Hypermedia and Multimedia 25(4) pp. 268-287. 10.1080/13614568.2020.1722761.
da Costa, Pablo Botton; Ivandré Paraboni (2019) Personality-dependent neural text summarization. Recent Advances in Natural Language Processing (RANLP-2019), pp. 205-212. Varna, Bulgaria. 10.26615/978-954-452-056-4_024.
Custódio, José Eleandro ; Ivandré Paraboni (2019) An Ensemble Approach to Cross-Domain Authorship Attribution. 10th Conference and Labs of the Evaluation Forum (CLEF-2019). Lugano, Switzerland. LNCS vol. 11696, pp. 201-212. Springer. 10.1007/978-3-030-28577-7_17.
Mariotti, André Costa; Ivandré Paraboni (2019) Generating Customizable Natural Language Descriptions. IEEE Latin America Transactions vol. 17 nr. 08, pp. 1252-1258. 10.1109/TLA.2019.8932333.
Custódio, José Eleandro ; Ivandré Paraboni (2019) Multichannel Open-set Cross-Domain Authorship Attribution. 10th Conference and Labs of the Evaluation Forum (CLEF-2019). Lugano, Switzerland. Working Notes Papers of the CLEF 2019 Evaluation Lab.
dos Santos, Wesley Ramos ; Ivandré Paraboni (2019) Personality facets recognition from text. 10th Conference and Labs of the Evaluation Forum (CLEF-2019). Lugano, Switzerland. LNCS vol. 11696, pp. 185-190. Springer. 10.1007/978-3-030-28577-7_15.
Dias, Rafael Felipe Sandroni ; Ivandré Paraboni (2019) Combined CNN+RNN Bot and Gender Profiling. 10th Conference and Labs of the Evaluation Forum (CLEF-2019). Lugano, Switzerland. Working Notes Papers of the CLEF 2019 Evaluation Lab.
Dias, Rafael Felipe Sandroni ; Ivandré Paraboni (2018) Caracterização autoral de usuários do Facebook: gênero, idade, religiosidade e área de formação. I Congresso Internacional em Humanidades Digitais (HDRio-2018), Rio de Janeiro, pp. 408-412.
Custódio, José Eleandro ; Ivandré Paraboni (2018) Similaridade de Textos aplicada à Verificação Autoral. I Congresso Internacional em Humanidades Digitais (HDRio-2018), Rio de Janeiro, pp. 403-407.
dos Santos, Wesley Ramos; Ivandré Paraboni (2018) Reconhecimento Automático de Traços de Personalidade em Redes Sociais. I Congresso Internacional em Humanidades Digitais (HDRio-2018), Rio de Janeiro, pp. 398-402.
Silva, Barbara B. C.; Ivandré Paraboni (2018) Personality Recognition from Facebook Text. 13th International Conference on the Computational Processing of Portuguese (PROPOR-2018), 2018, Canela. Lecture Notes in Computer Science, 2018. v. 11122. pp. 107-114. (selected among the top three finalists of the Best Dissertation on Language Technology for Portuguese Award 2015-2018)
Dias, Rafael Felipe Sandroni ; Ivandré Paraboni. Author Profiling using Word Embeddings with Subword Information. Conference and Labs of the Evaluation Forum (CLEF-2018), 2018, Avignon. Working Notes Papers of the CLEF 2018 Evaluation Labs, 2018. v. 2125.
Custódio, José Eleandro ; Ivandré Paraboni. EACH-USP Ensemble Cross-Domain Authorship Attribution. Conference and Labs of the Evaluation Forum (CLEF-2018), 2018, Avignon. Working Notes Papers of the CLEF 2018 Evaluation Lab, 2018. v. 2125. (best-performing participant system of the PAN-CLEF 2018 cross-domain authorship attribution shared task)
Silva, Barbara B. C.; Ivandré Paraboni (2018) Learning Personality Traits from Facebook Text. IEEE Latin America vol. 16 nr. 4. pp. 1256-1262. 10.1109/TLA.2018.8362165.
Ramos, Ricelli Moreira Silva; Georges Basile Stavracas Neto; Barbara Barbosa Claudino Silva; Danielle Sampaio Monteiro; Ivandré Paraboni; Rafael Felipe Sandroni Dias (2018) Building a Corpus for Personality-dependent Natural Language Understanding and Generation. 11th International Conference on Language Resources and Evaluation (LREC-2018) pp. 1138-1145. Miyazaki, Japan.
Hsieh, Fernando Chiu; Rafael Felipe Sandroni Dias; Ivandré Paraboni (2018) Author Profiling from Facebook Corpora. 11th International Conference on Language Resources and Evaluation (LREC-2018) pp. 2566-2570. Miyazaki, Japan.
Lan, Alex Gwo Jen; Ivandré Paraboni (2018) Definite Description Lexical Choice: taking speaker's Personality into account. 11th International Conference on Language Resources and Evaluation (LREC-2018) pp. 2999-3004. Miyazaki, Japan.
dos Santos, Vitor Garcia; Ivandré Paraboni; Bárbara Barbosa Claudino Silva (2017) Big Five Personality Recognition from Multiple Text Genres. Text, Speech and Dialogue (TSD-2017). LNAI vol.10415, pp. 29-37. Springer. 10.1007/978-3-319-64206-2_4. Prague, Czech Republic. The original publication is available at http://www.springerlink.com.
Paraboni, Ivandré; Danielle Sampaio Monteiro ; Alex Gwo Jen Lan (2017) Personality-dependent Referring Expression Generation. Text, Speech and Dialogue (TSD-2017). LNAI vol. 10415, pp. 20-28. Springer. 10.1007/978-3-319-64206-2_3. Prague, Czech Republic. The original publication is available at http://www.springerlink.com.