ValuesML: A new multilingual dataset for values detection in news and political manifestos

Mario Scharfbillig; Theresa Reitis-Münstermann; Nicolas Stefanovitsch; Johannes Kiesel; Paula Schulze Brock; Joanne Sneddon; Emmanuel Cartier; Murat Ardag; Sharon Arieli; Ella Daniel; Henrik Dobewall; Johannes Karl; Anna Krasteva; Thomas Peter Oeschger; Georgios Petasis; Luana Russo; Antonella Seddone; Aurelia Tamo-Larrieux; Hester van Herk; Nazan Avci; Giuliano Bobba; Pierre Chiron; Ahmet Coymak; Maria Dagioglou; Dara Katsamori; Ingmar Leijen; Berna Öney; Ricarda Scholz-Kuhn; Emilia Zankina; Sandrine Astor; Petra Auer; Livia Aulino; Anat Bardi; Fiorella Battaglia; Constanze Beierlein; Maya Benish-Weisman; Christina Christodoulou; Patricia Collins; Irene Coppola; Mafalda Damaso; Meike Morren; Einat Elizarov; Naama Erlich; Peculiar Ezeigwe-Ephraim; Ronald Fischer; Maria Christina Gaeta; Lucilla Gatt; Noam Gerera; Sjoukje Goldman; Frederic Gonthier; Stefanie Habermann; Mina Hristova; Demet Islambay Yapali; Luigi Izzo; Panos Kapetanakis; Resit Kislioglu; Chaya Koleva; Roberta Koleva; Josua Lake; Mael Mesplou; Vanina Ninova; Elif Sandal Önal; Shani Oppenheim-Weller; Duygu Ozturk; Ioannis Ellissaios Paparrigopoulos; Vladimir Ponizovskiy; Tim Reeskens; Maria Fancesca Romano; Torven Schalk; Alina Starovolsky-Shitrit; Oscar Smallenbroek; Ömer Topuz; Christin-Melanie Vauclair; Giuseppe Di Vetta; Sheng Ye

doi:10.3758/s13428-026-03092-z

ValuesML: A new multilingual dataset for values detection in news and political manifestos

Mario Scharfbillig^*
, Theresa Reitis-Münstermann
, Nicolas Stefanovitsch
, Johannes Kiesel
, Paula Schulze Brock
, Joanne Sneddon
, Emmanuel Cartier
, Murat Ardag
, Sharon Arieli
, Ella Daniel
, Henrik Dobewall
, Johannes Karl
, Anna Krasteva
, Thomas Peter Oeschger
, Georgios Petasis
, Luana Russo
, Antonella Seddone
, Aurelia Tamo-Larrieux
, Hester van Herk
, Nazan Avci

Giuliano Bobba, Pierre Chiron, Ahmet Coymak, Maria Dagioglou, Dara Katsamori, Ingmar Leijen, Berna Öney, Ricarda Scholz-Kuhn, Emilia Zankina, Sandrine Astor, Petra Auer, Livia Aulino, Anat Bardi, Fiorella Battaglia, Constanze Beierlein, Maya Benish-Weisman, Christina Christodoulou, Patricia Collins, Irene Coppola, Mafalda Damaso, Meike Morren, Einat Elizarov, Naama Erlich, Peculiar Ezeigwe-Ephraim, Ronald Fischer, Maria Christina Gaeta, Lucilla Gatt, Noam Gerera, Sjoukje Goldman, Frederic Gonthier, Stefanie Habermann, Mina Hristova, Demet Islambay Yapali, Luigi Izzo, Panos Kapetanakis, Resit Kislioglu, Chaya Koleva, Roberta Koleva, Josua Lake, Mael Mesplou, Vanina Ninova, Elif Sandal Önal, Shani Oppenheim-Weller, Duygu Ozturk, Ioannis Ellissaios Paparrigopoulos, Vladimir Ponizovskiy, Tim Reeskens, Maria Fancesca Romano, Torven Schalk, Alina Starovolsky-Shitrit, Oscar Smallenbroek, Ömer Topuz, Christin-Melanie Vauclair, Giuseppe Di Vetta, Sheng Ye

^*Corresponding author for this work

Sociology

Research output: Contribution to journal › Article › Scientific › peer-review

Abstract

Values are important building blocks of political ideologies and are frequently invoked in political debates. Yet, because values are abstract, understanding how they are expressed in real-world discourse, and how this varies across contexts, remains a challenge. In this paper, we introduce a large-scale, expert-annotated dataset of value expression in political text, comprising news articles and political manifestos across nine languages. The dataset is grounded in the refined theory of human values (Schwartz et al., 2012) and was developed through an iterative process of annotation guideline development, annotator training, and expert curation. Value annotations capture both the type of value expressed and its evaluative framing, distinguishing whether values are expressed as attained or constrained. The final dataset comprises 2648 texts and 74,231 sentences across nine languages. This resource enables systematic, cross-linguistic analysis of value expression in political communication and provides a benchmark for developing and evaluating computational models of value detection.

Original language	English
Article number	276
Journal	Behavior Research Methods
Volume	58
DOIs	https://doi.org/10.3758/s13428-026-03092-z
Publication status	Published - 20 Aug 2026

Access to Document

10.3758/s13428-026-03092-z

Cite this

Scharfbillig, M., Reitis-Münstermann, T., Stefanovitsch, N., Kiesel, J., Schulze Brock, P., Sneddon, J., Cartier, E., Ardag, M., Arieli, S., Daniel, E., Dobewall, H., Karl, J., Krasteva, A., Oeschger, T. P., Petasis, G., Russo, L., Seddone, A., Tamo-Larrieux, A., van Herk, H., ... Ye, S. (2026). ValuesML: A new multilingual dataset for values detection in news and political manifestos. Behavior Research Methods, 58, Article 276. https://doi.org/10.3758/s13428-026-03092-z

@article{951bd8c8155048e69f4e20e234b9cfb5,

title = "ValuesML: A new multilingual dataset for values detection in news and political manifestos",

abstract = "Values are important building blocks of political ideologies and are frequently invoked in political debates. Yet, because values are abstract, understanding how they are expressed in real-world discourse, and how this varies across contexts, remains a challenge. In this paper, we introduce a large-scale, expert-annotated dataset of value expression in political text, comprising news articles and political manifestos across nine languages. The dataset is grounded in the refined theory of human values (Schwartz et al., 2012) and was developed through an iterative process of annotation guideline development, annotator training, and expert curation. Value annotations capture both the type of value expressed and its evaluative framing, distinguishing whether values are expressed as attained or constrained. The final dataset comprises 2648 texts and 74,231 sentences across nine languages. This resource enables systematic, cross-linguistic analysis of value expression in political communication and provides a benchmark for developing and evaluating computational models of value detection.",

author = "Mario Scharfbillig and Theresa Reitis-M{\"u}nstermann and Nicolas Stefanovitsch and Johannes Kiesel and \{Schulze Brock\}, Paula and Joanne Sneddon and Emmanuel Cartier and Murat Ardag and Sharon Arieli and Ella Daniel and Henrik Dobewall and Johannes Karl and Anna Krasteva and Oeschger, \{Thomas Peter\} and Georgios Petasis and Luana Russo and Antonella Seddone and Aurelia Tamo-Larrieux and \{van Herk\}, Hester and Nazan Avci and Giuliano Bobba and Pierre Chiron and Ahmet Coymak and Maria Dagioglou and Dara Katsamori and Ingmar Leijen and Berna {\"O}ney and Ricarda Scholz-Kuhn and Emilia Zankina and Sandrine Astor and Petra Auer and Livia Aulino and Anat Bardi and Fiorella Battaglia and Constanze Beierlein and Maya Benish-Weisman and Christina Christodoulou and Patricia Collins and Irene Coppola and Mafalda Damaso and Meike Morren and Einat Elizarov and Naama Erlich and Peculiar Ezeigwe-Ephraim and Ronald Fischer and Gaeta, \{Maria Christina\} and Lucilla Gatt and Noam Gerera and Sjoukje Goldman and Frederic Gonthier and Stefanie Habermann and Mina Hristova and \{Islambay Yapali\}, Demet and Luigi Izzo and Panos Kapetanakis and Resit Kislioglu and Chaya Koleva and Roberta Koleva and Josua Lake and Mael Mesplou and Vanina Ninova and {\"O}nal, \{Elif Sandal\} and Shani Oppenheim-Weller and Duygu Ozturk and Paparrigopoulos, \{Ioannis Ellissaios\} and Vladimir Ponizovskiy and Tim Reeskens and Romano, \{Maria Fancesca\} and Torven Schalk and Alina Starovolsky-Shitrit and Oscar Smallenbroek and {\"O}mer Topuz and Christin-Melanie Vauclair and \{Di Vetta\}, Giuseppe and Sheng Ye",

year = "2026",

month = aug,

day = "20",

doi = "10.3758/s13428-026-03092-z",

language = "English",

volume = "58",

journal = "Behavior Research Methods",

issn = "1554-351X",

publisher = "Springer Nature",

}

Scharfbillig, M, Reitis-Münstermann, T, Stefanovitsch, N, Kiesel, J, Schulze Brock, P, Sneddon, J, Cartier, E, Ardag, M, Arieli, S, Daniel, E, Dobewall, H, Karl, J, Krasteva, A, Oeschger, TP, Petasis, G, Russo, L, Seddone, A, Tamo-Larrieux, A, van Herk, H, Avci, N, Bobba, G, Chiron, P, Coymak, A, Dagioglou, M, Katsamori, D, Leijen, I, Öney, B, Scholz-Kuhn, R, Zankina, E, Astor, S, Auer, P, Aulino, L, Bardi, A, Battaglia, F, Beierlein, C, Benish-Weisman, M, Christodoulou, C, Collins, P, Coppola, I, Damaso, M, Morren, M, Elizarov, E, Erlich, N, Ezeigwe-Ephraim, P, Fischer, R, Gaeta, MC, Gatt, L, Gerera, N, Goldman, S, Gonthier, F, Habermann, S, Hristova, M, Islambay Yapali, D, Izzo, L, Kapetanakis, P, Kislioglu, R, Koleva, C, Koleva, R, Lake, J, Mesplou, M, Ninova, V, Önal, ES, Oppenheim-Weller, S, Ozturk, D, Paparrigopoulos, IE, Ponizovskiy, V, Reeskens, T, Romano, MF, Schalk, T, Starovolsky-Shitrit, A, Smallenbroek, O, Topuz, Ö, Vauclair, C-M, Di Vetta, G & Ye, S 2026, 'ValuesML: A new multilingual dataset for values detection in news and political manifestos', Behavior Research Methods, vol. 58, 276. https://doi.org/10.3758/s13428-026-03092-z

TY - JOUR

T1 - ValuesML: A new multilingual dataset for values detection in news and political manifestos

AU - Scharfbillig, Mario

AU - Reitis-Münstermann, Theresa

AU - Stefanovitsch, Nicolas

AU - Kiesel, Johannes

AU - Schulze Brock, Paula

AU - Sneddon, Joanne

AU - Cartier, Emmanuel

AU - Ardag, Murat

AU - Arieli, Sharon

AU - Daniel, Ella

AU - Dobewall, Henrik

AU - Karl, Johannes

AU - Krasteva, Anna

AU - Oeschger, Thomas Peter

AU - Petasis, Georgios

AU - Russo, Luana

AU - Seddone, Antonella

AU - Tamo-Larrieux, Aurelia

AU - van Herk, Hester

AU - Avci, Nazan

AU - Bobba, Giuliano

AU - Chiron, Pierre

AU - Coymak, Ahmet

AU - Dagioglou, Maria

AU - Katsamori, Dara

AU - Leijen, Ingmar

AU - Öney, Berna

AU - Scholz-Kuhn, Ricarda

AU - Zankina, Emilia

AU - Astor, Sandrine

AU - Auer, Petra

AU - Aulino, Livia

AU - Bardi, Anat

AU - Battaglia, Fiorella

AU - Beierlein, Constanze

AU - Benish-Weisman, Maya

AU - Christodoulou, Christina

AU - Collins, Patricia

AU - Coppola, Irene

AU - Damaso, Mafalda

AU - Morren, Meike

AU - Elizarov, Einat

AU - Erlich, Naama

AU - Ezeigwe-Ephraim, Peculiar

AU - Fischer, Ronald

AU - Gaeta, Maria Christina

AU - Gatt, Lucilla

AU - Gerera, Noam

AU - Goldman, Sjoukje

AU - Gonthier, Frederic

AU - Habermann, Stefanie

AU - Hristova, Mina

AU - Islambay Yapali, Demet

AU - Izzo, Luigi

AU - Kapetanakis, Panos

AU - Kislioglu, Resit

AU - Koleva, Chaya

AU - Koleva, Roberta

AU - Lake, Josua

AU - Mesplou, Mael

AU - Ninova, Vanina

AU - Önal, Elif Sandal

AU - Oppenheim-Weller, Shani

AU - Ozturk, Duygu

AU - Paparrigopoulos, Ioannis Ellissaios

AU - Ponizovskiy, Vladimir

AU - Reeskens, Tim

AU - Romano, Maria Fancesca

AU - Schalk, Torven

AU - Starovolsky-Shitrit, Alina

AU - Smallenbroek, Oscar

AU - Topuz, Ömer

AU - Vauclair, Christin-Melanie

AU - Di Vetta, Giuseppe

AU - Ye, Sheng

PY - 2026/8/20

Y1 - 2026/8/20

N2 - Values are important building blocks of political ideologies and are frequently invoked in political debates. Yet, because values are abstract, understanding how they are expressed in real-world discourse, and how this varies across contexts, remains a challenge. In this paper, we introduce a large-scale, expert-annotated dataset of value expression in political text, comprising news articles and political manifestos across nine languages. The dataset is grounded in the refined theory of human values (Schwartz et al., 2012) and was developed through an iterative process of annotation guideline development, annotator training, and expert curation. Value annotations capture both the type of value expressed and its evaluative framing, distinguishing whether values are expressed as attained or constrained. The final dataset comprises 2648 texts and 74,231 sentences across nine languages. This resource enables systematic, cross-linguistic analysis of value expression in political communication and provides a benchmark for developing and evaluating computational models of value detection.

AB - Values are important building blocks of political ideologies and are frequently invoked in political debates. Yet, because values are abstract, understanding how they are expressed in real-world discourse, and how this varies across contexts, remains a challenge. In this paper, we introduce a large-scale, expert-annotated dataset of value expression in political text, comprising news articles and political manifestos across nine languages. The dataset is grounded in the refined theory of human values (Schwartz et al., 2012) and was developed through an iterative process of annotation guideline development, annotator training, and expert curation. Value annotations capture both the type of value expressed and its evaluative framing, distinguishing whether values are expressed as attained or constrained. The final dataset comprises 2648 texts and 74,231 sentences across nine languages. This resource enables systematic, cross-linguistic analysis of value expression in political communication and provides a benchmark for developing and evaluating computational models of value detection.

U2 - 10.3758/s13428-026-03092-z

DO - 10.3758/s13428-026-03092-z

M3 - Article

SN - 1554-351X

VL - 58

JO - Behavior Research Methods

JF - Behavior Research Methods

M1 - 276

ER -

ValuesML: A new multilingual dataset for values detection in news and political manifestos

Abstract

Access to Document

Fingerprint

Cite this