Cross-Media Sentiment Analysis on German Blogs

  • Nina N. Zahn Universidade de Mannheim
  • Greice P. Dal Molin PUCRS
  • Soraia R. Musse PUCRS


Social interactions have changed in recent years. People post their thoughts, opinions and feelings on social media platforms more often. Due to the increase in the amount of data on the internet, it is impracticable to carry out the sentiment analysis manually, requiring automation of the process. In this work, we present the corpus Cross-Media German Blog (CGB) which consists of German blogs with feelings in the domain of images, texts and posts (Ground Truth), classified according to human perceptions. We apply existing Machine Learning technologies and lexicons to the corpus to detect the feelings (negative, neutral or positive) of the images and texts and compare the results with the GT. We examined contradictory posts, when the image and text classified by humans in the same post had diverging feelings. The comparison of this article with the analysis of sentiment among the media of Brazilian blogs finds its justification for performance results in cultural differences, since, throughout this work, Brazil is classified as indulgent and Germany as a restrained country.
Palavras-chave: cross-media sentiment analysis, corpus, emotions in images and texts, cultural differences


Borth, D., Chen, T., Ji, R., and Chang, S.-F. (2013). Sentibank: large-scale ontology andclassifiers for detecting sentiment and emotions in visual content. InProceedings ofthe 21st ACM international conference on Multimedia, pages 459–460

Camras, L. (1980). Emotion: a psychoevolutionary synthesis

Chen, T., Borth, D., Darrell, T., and Chang, S.-F. (2014). Deepsentibank: Visual senti-ment concept classification with deep convolutional neural networks.arXiv preprintarXiv:1410.8586

Dal Molin, G. P., Santos, H. D., Manssour, I. H., Vieira, R., and Musse, S. R. (2019).Cross-media sentiment analysis in brazilian blogs. InInternational Symposium onVisual Computing, pages 492–503. Springer.

De Marneffe, M.-C., Rafferty, A. N., and Manning, C. D. (2008). Finding contradictionsin text. InProceedings of ACL-08: HLT, pages 1039–1047.

Hamp, B. and Feldweg, H. (1997). Germanet-a lexical-semantic net for german. InAutomatic information extraction and building of lexical semantic resources for NLPapplications

Henrich, V. and Hinrichs, E. (2010). Gernedit-the germanet editing tool. InProceedingsof the ACL 2010 System Demonstrations, pages 19–24.

Hofstede, G. (2001).Culture’s consequences: Comparing values, behaviors, institutionsand organizations across nations. Sage publications.

Hofstede, G. (2011). Dimensionalizing cultures: The hofstede model in context.Onlinereadings in psychology and culture, 2(1):2307–0919

slam, J. and Zhang, Y. (2016). Visual sentiment analysis for social images using transferlearning approach. In2016 IEEE International Conferences on Big Data and CloudComputing (BDCloud), Social Computing and Networking (SocialCom), SustainableComputing and Communications (SustainCom)(BDCloud-SocialCom-SustainCom),pages 124–130. IEEE.

Meier, T., Boyd, R. L., Pennebaker, J. W., Mehl, M. R., Martin, M., Wolf, M., and Horn,A. B. (2019). “liwc auf deutsch”: The development, psychometrics, and introductionof de-liwc2015.PsyArXiv, (a)

Moraes, S. M., Santos, A. L., Redecker, M., Machado, R. M., and Meneguzzi, F. R.(2016). Comparing approaches to subjectivity classification: A study on portuguesetweets. InInternational Conference on Computational Processing of the PortugueseLanguage, pages 86–94. Springer

Morency, L.-P., Mihalcea, R., and Doshi, P. (2011). Towards multimodal sentiment analy-sis: Harvesting opinions from the web. InProceedings of the 13th international con-ference on multimodal interfaces, pages 169–176

Rauh, C. (2018). Validating a sentiment dictionary for german political language—aworkbench note.Journal of Information Technology & Politics, 15(4):319–343.

Soleymani, M., Garcia, D., Jou, B., Schuller, B., Chang, S.-F., and Pantic, M. (2017). Asurvey of multimodal sentiment analysis.Image and Vision Computing, 65:3–14.

Tymann, K., Lutz, M., Palsbr ̈oker, P., and Gips, C. (2019). Gervader-a german adaptationof the vader sentiment analysis tool for social media texts. InLWDA, pages 178–189

Vadicamo, L., Carrara, F., Cimino, A., Cresci, S., Dell’Orletta, F., Falchi, F., and Tesconi,M. (2017). Cross-media learning for image sentiment analysis in the wild. In2017IEEE International Conference on Computer Vision Workshops (ICCVW), pages 308–317

Vinodhini, G. and Chandrasekaran, R. (2012). Sentiment analysis and opinion mining: asurvey.International Journal, 2(6):282–292

Waltinger, U. (2010). Germanpolarityclues: A lexical resource for german sentimentanalysis. InLREC, pages 1638–1642

Wartena, C. (2019). A probabilistic morphology model for german lemmatization

Zadeh, A., Chen, M., Poria, S., Cambria, E., and Morency, L.-P. (2017). Tensor fusionnetwork for multimodal sentiment analysis.arXiv preprint arXiv:1707.07250
ZAHN, Nina N.; DAL MOLIN, Greice P.; MUSSE, Soraia R.. Cross-Media Sentiment Analysis on German Blogs. In: SEMINÁRIO INTEGRADO DE SOFTWARE E HARDWARE (SEMISH), 48. , 2021, Evento Online. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2021 . p. 114-122. ISSN 2595-6205. DOI: