October 11, 2024 | LMU Climate School

As in the previous year, the KLIMA-MEMES project was once again part of the LMU Climate School in 2024. At the interdisciplinary event on October 11, Simon Lübke led a workshop on news media and climate change. In the workshop, participants discussed the basics of journalistic reporting on climate change. They learned how to recognize and identify reporting patterns in climate change reporting. The workshop also provided insights into current trends in the climate debate were discussed, including constructive journalism and the role of internet memes and humor.

October 3, 2024 | ECCV 2024

Johannes Schusterbauer and Björn Ommer took part in the 8th European Conference on Computer Vision ECCV 2024 in Milan from September 29 to October 4. Among other things, they presented the project paper entitled “FMBoost: Boosting Latent Diffusion with Flow Matching” and were selected for an oral presentation. In this paper, they present a method that can be used to achieve complex image processing models with less computational effort.

Abstract: Visual synthesis has recently seen significant leaps in performance, inparticular due to breakthroughs in generative models. Diffusion models have been a key enabler as they excel in image diversity. This, however, comes at the prize of slow training and synthesis which are only partially alleviated by latent diffusion. To this end, flow matching is an appealing approach due to its complementary characteristics of faster training and inference but less diverse synthesis. We demonstrate that introducing flow matching between a frozen diffusion model and convolutional decoder enables high-resolution image synthesis at reduced computational cost and model size. A small diffusion model can then provide the necessary visual diversity effectively, while flow matching efficiently enhances resolution and details by mapping the small to a high-dimensional latent space. These latents are then projected to high-resolution images by the subsequent convolutional decoder of the diffusion approach. Combining the diversity of diffusion models, the efficiency of flow matching, and the effectiveness of convolutional decoders, achieves state-of-the-art high-resolution image synthesis at 1024 ² pixels with minimal computational cost. Cascading our model optionally boosts this further to 2048 ² pixels. Importantly, our approach is orthogonal to recent approximation and speed-up strategies for the underlying model, making it easily integrable into the various diffusion model frameworks.

September 25, 2024 | ECREA 2024 Conference

Jörg Haßler and Simon Lübke attended the 10th ECREA European Communication Conference in Ljubljana last week, from September 24 to 27. During the conference, Jörg Haßler presented their joint work with Simon Lübke, Nadezdha Ozornina, and Mario Haim titled „The use of humor styles in memes on climate change“ in the panel „Public understandings of climate change.“ The presentation discussed various humor styles used in internet memes and shared findings from our project’s analyses on the use of humor in content related to COP28.

September 20, 2024 | forsa / Paul Lazarsfeld Scholarship for Nadezdha Ozornina

Nadezdha Ozornina has been awarded the forsa / Paul Lazarsfeld Scholarship 2024 by the Methods Division of the German Communication Association (DGPuK). Our colleague from the KLIMA-MEMES project was honored with this award for her work „Comparing Multilingual Topic Models“ at the joint annual conference of the Methods and Digital Communication Divisions of the DGPuK in Hamburg. The Paul Lazarsfeld Scholarship recognizes outstanding students or graduates in journalism and communication studies for particularly demanding or innovative research and final theses in the field of methods. The entire project team congratulates Nadezdha on this achievement.

September 19, 2024 | Interview with Jörg Haßler on the political communication by German parties on TikTok

The political communication of the Alternative für Deutschland (AfD) on social media is often cited as one of the reasons for the party’s recent electoral successes. Specifically, the party’s content on TikTok is highlighted to explain the high voter turnout among younger demographics. In an interview with Deutschlandfunk, the project leader of KLIMA-MEMES, Jörg Haßler, analyzes the AfD’s communication strategy on TikTok. He points to the heightened emotionalization of AfD content as a key factor explaining the higher average reach of AfD posts compared to those of other parties. These and other insights from Jörg Haßler can be found in the interview titled “ „Soziale Medien – Wie demokratische Parteien auf TikTok nachziehen“ by Tobias Krone.

September 19, 2024 | Conference of the DGPuK Methods Division

Nadja Ozornina took part in the joint annual conference of the DGPuK Methods and Digital Communication Divisions in Hamburg from September 18 to 20. As part of the conference, she presented a joint paper with Simon Lübke and Mario Haim and entitled “More than the sum of its parts: Challenges and approaches of automated analysis of multimodal social media content”. The paper deals with the analysis of multimodal social media content, i.e. content that combines multiple modalities such as text and image to construct meaning. The presentation discussed the methodological challenges that exist when analyzing multimodal content and presented various approaches to solving them.

Abstract: In the context of datafication, social media posts are an important source for automated analyses in communication science. Multimodal content, which combines different modalities such as text, images and sound, is playing an increasingly important role. However, analyzing this content poses methodological challenges for researchers, as images and texts have to be processed differently and it is often unclear how the information from the different modalities can be linked. Current research methods in communication science are mostly unimodal, while an overview of existing multimodal approaches and their characteristics is lacking. The article identifies three different groups of approaches for analyzing multimodal content, which differ in terms of the number of concepts examined and the respective intermediate analytical steps. Various application examples for the three approaches are presented and the advantages and disadvantages of the individual methods for different questions are discussed.

September 16, 2024 | Interview on pro-Russian cartoons and memes

In an investigative report by NDR, WDR and Süddeutsche Zeitung, internal documents from the Moscow-based Social Design Agency (SDA) were analyzed, providing insight into Russian disinformation campaigns in Germany. The documents show that Germany is one of the Kremlin’s preferred targets for spreading disinformation. In this context, the SDA disseminates numerous cartoons and internet memes on social media that specifically convey fake news and pro-Russian narratives. Our colleague from the KLIMA-MEMES project, Simon Lübke, analyzed some of the cartoons from the data leak for the Süddeutsche Zeitung. He also provides a short assessment of the conditions under which this content can be effective.

The article with the research results and Lübke’s statements can be accessed via the following link.

August 16, 2024 | ACL 2024 Conference

Shijia Zhou, Siyao Peng, and Barbara Plank from the KLIMA-MEMES Project attended the 62nd Annual Meeting of the Association for Computational Lunguistics (ACL 2024) from August 11-16 in Bangkok to present results from several research projects. Among them was a project presentation entitled “CLIMATELI: Evaluating Entity Linking on Climate Change Data” in which they use the entity linking approach to understand who and what is mentioned in different climate change text content.

Abstract: Climate Change (CC) is a pressing topic of global importance, attracting increasing attention across research fields, from social sciences to Natural Language Processing (NLP). CC is also discussed in various settings and communication platforms, from academic publications to social media forums. Understanding who and what is mentioned in such data is a first critical step to gaining new insights into CC. We present CLIMATELI (CLIMATe Entity LInking), the first manually annotated CC dataset that links 3,087 entity spans to Wikipedia. Using CLIMATELI (CLIMATe Entity LInking), we evaluate existing entity linking (EL) systems on the CC topic across various genres and propose automated filtering methods for CC entities. We find that the performance of EL models notably lags behind humans at both token and entity levels. Testing within the scope of retaining or excluding non-nominal and/or non-CC entities particularly impacts the models‘ performances.

June 25, 2024 |Talk by Manfred Stede

Prof. Manfred Stede, Professor of Applied Computational Linguistics at the University of Potsdam, visited Barbara Plank’s MaiNLP research lab last week. During his visit, he gave a talk on the current projects of his research group titled „NLP on Climate Change Discourse: Two Case Studies.“ The talk focused on two main topics: the framing of climate change in journal editorials and the analysis of communication by members of the German Bundestag.

Abstract: The debate around climate change (CC) — its extent, its causes, and the necessary responses — is intense and of global importance. The ongoing discourses are a prominent object of study in several Social Sciences, while in the natural language processing community, this domain has so far received relatively little attention. In my talk, I first give a brief overview of types of approaches and data, and then report on two case studies that we are currently conducting in my research group. The first tackles the notion of „framing“ (the perspective taken in viewing an issue) in CC-related editorials of the journals ‚Nature‘ and ‚Science‘: We proceed from a coarse-grained text-level labeling to increasingly detailed clause-level annotation of framing CC, and run experiments on automatic classification. The second involves a corpus of parliamentary speeches, press releases and tweets from the members of the German parliament (2017-2021) and compares their ways of addressing CC, contrasting on the one hand the different communication channels and on the other hand the party affiliations of the speakers.

Bio: I’m a professor of applied computational linguistics at Potsdam University. My research and teaching activities revolve around issues in discourse structure and automatic text understanding, currently often for research questions from the social sciences.

June 21, 2024 | Blog entry on the use humor in the climate crises

During re:publica 24, the KLIMA-MEMES team presented initial results from the research project. Simon Lübke has now also summarized these results and the contents of the presentation in a blog post for the Bavarian Research Institute for Digital Transformation (bid). The blog post, titled „The Role of Humor in the Climate Crisis,“ discusses the function and impact of memes and humorous content in political communication. Additionally, the first findings of the manual content analysis of the COP28 data conducted by the research team are presented and contextualized. Please find the link to the post below.

May 28, 2024 | re:publica24

From May 27-29, 2024, Simon Lübke participated in re:publica24 as a speaker. As part of the festival for the digital society, he gave a talk titled “Humor in the Climate Crisis: How Memes Influence the Political Climate Discourse.” In this presentation, he shared insights and first findings from the Climate Memes Project. These included preliminary results from a manual content analysis on the use of humor and memes during the 2023 UN Climate Change Conference (COP28) in Dubai.

Abstract: Memes are considered an effective strategic format in political discourses. However, it is still unclear what role they play in the climate change debate and which actors are actively involved. With the KLIMA-MEMES project, we investigate the use of climate change memes on social media, focusing on which memes are particularly popular and why. Recently, we analyzed about 900,000 social media posts on Instagram, Facebook, TikTok, and YouTube during the 2023 UN Climate Change Conference #COP28 in Dubai. Our initial findings show that memes achieve significant reach, especially when they use negative humor. However, relevant political actors such as protest groups and politicians have so far rarely used humorous content and memes. Thus, they miss the opportunity to counter popular and aggressive climate crisis memes with positive, humorous content.

May 3, 2024 | COMPTEXT Conference

On 2-4 May 2024, Nadezhda Ozornina participated in ‚The Sixth International and Interdisciplinary Conference on the Quantitative and Computational Analysis of Textual Data (COMPTEXT)‘ in Amsterdam, the Netherlands. She presented her study on the application of machine translation for multilingual topic modeling which investigates how robust the results of automated text analysis are depending on translation in different languages. The implications of the study provide a basis for analyzing data on climate change from social media and serve as recommendations for future studies. Additionally, there were meetings with other multimodal researchers that offered opportunities for interdisciplinary collaboration.

Abstract: For the consolidation of multilingual texts for topic modeling, machine translation is one of the most prominent approaches. However, when applying machine translation, the impacts of target language choice on the outcomes of topic modeling remain unclear. This study tests these impacts via two strategies: (a) consolidating texts into one of the documents’ original languages and (b) translating texts into an intermediary language not present in the initial data. To compare possible outcomes, a corpus of parallel texts from the United Nations written in Russian and German (N = 3760) is examined. The entire dataset is consolidated into one of the original languages by translating Russian texts into German, followed by the application of structural topic modeling algorithms. The results are then compared with outcomes derived from topic modeling after translating the same data into an intermediary language (English). For each translation approach, the similarity between parallel text collections is measured using topical prevalence and topical content metrics. The findings indicate that using an intermediary language for machine translation leads to a more symmetrical distribution of topics and a higher percentage of overlapping top words compared to consolidating the texts into one of the original languages. However, translating into an intermediary language significantly reduces the vocabulary as compared to retaining some texts in their original language. The findings are validated on different numbers of topics. Based on the results, the implications of using different target languages for machine translation are discussed and recommendations for future studies are developed.

April 30, 2024 | Collecting and Analyzing Social Media Data

Nadja Ozornina and Simon Lübke presented recent findings from the Climate Memes project during a seminar held by the MaiNLP research lab at LMU. The seminar, led by KLIMA-MEMES team member Siyao (Logan) Peng, focused on Natural Language Processing (NLP) for Climate Change. The main focus of the talk titled „Collecting and Analyzing Social Media Data for Interdisciplinary Research” was on various approaches to collecting social media data from different platforms. Additionally, the two researchers engaged with the course participants in discussions about the project’s findings on COP28 and explored various analysis strategies.

April 17, 2024 | Talk by Nadja Ozornina

Nadja Ozornina presented her work titled „When Multimodal Meets Comparative: Analyzing Multimodal Data in Comparative Context for Computational Communication Research“ to the KLIMA-MEMES team today. Her work aims to develop a methodological approach for computational communication research, which allows for multimodal analyses in comparative contexts. This includes, for example, the combined automated analysis of image and text content in social media posts in various languages.

Since October 2023, Nadja Ozornina has been a doctoral student at the Institute of Communication Science Munich and in the Computational Communication Research department. She also works as a research associate in the KLIMA-MEMES project. Before joining the project, she studied at the Higher School of Economics in Moscow and at LMU Munich.

February 16, 2024 | Challenges and Approaches in Interdisciplinary Research

Researchers from the Klima-Memes project took part in a networking meeting at the University of Vienna on February 15 and 16. As part of the meeting with colleagues from the Computational Communication Science Lab, Mario Haim, Nadja Ozornina and Simon Lübke presented initial results from the Klima-Memes project. The talk also included considerations on the challenges and approaches in interdisciplinary research projects. We would like to thank Prof. Annie Waldherr and Prof. Hajo Boomgaarden for the organization and their hospitality.

February 9, 2024 | DGPuK Political Communication Conference

Simon Lübke, Mario Haim and Jörg Haßler presented findings from the KLIMA-MEMES project at the annual conference of the „Communication and Politics“ section of of the German Communicaiton Association (DGPuK). The presentation was entitled „Climate communication multimodal: The role and effect of political internet memes in the climate discourse“. The conference took place from February 7 to 9 at the FH Graubünden in Bern on the topic of „Politics Multimedal“. The conference was jointly organized by the „Communication and Politics“ section of the  DGPuK, the „Politics and Communication“ working group of the German Political Science Association (DVPW) and the „Political Communication“ section of the Swiss Society for Communication and Media Studies (SGKM).

December 21, 2023 | Data Collection and COP28 Reports

From November 30 to December 13, the UN Climate Conference COP28 took place in Dubai. As part of the KLIMA-MEMES project, we collected social media posts on Facebook, Instagram, TikTok, and YouTube during the conference. Social media posts were included in our database if they used hashtags related to the climate conference, such as #COP28, or topics like climate change, for example, #climatechange. Additionally, we collected posts published by relevant actors in the German climate discourse. While we collected the data, we conducted first analyses of the social media communication during COP28 and published the our initial results in three reports. The reports focus on a comparison of the different social media platforms (Report 1), communication from relevant actor profiles (Report 2), and humorous content related to climate change, such as Political Internet Memes.

The reports can be read on this website under the section „COP28“.

November 27, 2023 | Research visit and CAS Lecture by Prof. Betsi Grabe

Professor Maria Elizabeth (Betsi) Grabe from Indiana University Bloomington visited the Department of Media and Communication Research in November at the invitation of the project team. As a leading researcher in the field of visual communication, Prof. Grabe serves as an international expert for the KLIMA-MEMES project. During her stay, she provided valuable guidance to colleagues involved in the KLIMA-MEMES project, particularly in the planning of the analysis of climate change memes. Furthermore, on November 27th, as part of the Climate Futures lecture series organized by the Center for Advanced Studies at LMU, Prof. Grabe had a talk titled „The Implications of Visual Primacy For Climate Messages.“

Abstract: Neuro and cognitive scientists have built the case for visual primacy over other forms of communication and meaning making by documenting (1) specialized centers in the human brain dedicated to visual processing (2) the speed with which visuals are processed, (3) the efficiency with which visual stimuli are classified and stored in memory, and (4) the way the visual system engages both emotional and “thinking” centers in the brain. This talk will present a review of the power of visuals on the one hand and the difficulties of effectively employing them in climate messages.

November 20, 2023 | Research visit and talk by Dr. Saif M. Mohammad

Dr. Saif M. Mohammad, Senior Research Scientist from the National Research Council Canada (NRC) and an external consultant for the KLIMA-MEMES project, recently visited Prof. Barbara Plank and the MaiNLP research lab in November. During his visit, he engaged with project members in discussions about the ongoing progress of the KLIMA-MEMES project and the manual annotation process of climate memes. On November 20, Dr. Mohammad, alongside Jan Philip Whale from the University of Göttingen, presented their collaborative project titled „We are Who We Cite: Bridges of Influence Between Natural Language Processing and Other Academic Fields“ at LMU.

Abstract: Natural Language Processing (NLP) is poised to substantially influence the world. However, significant progress comes hand-in-hand with substantial risks. Addressing them requires broad engagement with various fields of study. Yet, little empirical work examines the state of such engagement (past or current). In this paper, we quantify the degree of influence between 23 fields of study and NLP (on each other). We analyzed ~77k NLP papers, ~3.1m citations from NLP papers to other papers, and ~1.8m citations from other papers to NLP papers. We show that, unlike most fields, the cross-field engagement of NLP, measured by our proposed Citation Field Diversity Index (CFDI), has declined from 0.58 in 1980 to 0.31 in 2022 (an all-time low). In addition, we find that NLP has grown more insular — citing increasingly more NLP papers and having fewer papers that act as bridges between fields. NLP citations are dominated by computer science; Less than 8% of NLP citations are to linguistics, and less than 3% are to math and psychology. These findings underscore NLP’s urgent need to reflect on its engagement with various fields.

October 13, 2023  | Workshop Munich Climate School

From October 9 to October 13, 2023, the third Munich Climate School took place. At the Munich Climate School of LMU, students and educators from all disciplines come together to network. The goal of this interdisciplinary educational event is to comprehensively understand climate change in order to take proactive measures against it. As part of the event, Simon Lübke and Jörg Haßler from the KLIMA-MEMES team conducted a workshop titled „(News) Media and Climate Change.“

Abstract: The workshop provides an overview of the crucial role that (news) media play in the public discourse on climate change. Participants will learn about the fundamental processes of journalistic news reporting on climate change and understand the influence of news media in shaping public opinion and political actions on climate change. Based on current examples from the German climate discourse, we will explore how different news media report about climate change and discuss relevant factors that explain differences in the frequency and tonality of climate change reporting. We further discuss how political actors can gain public visibility by adapting to these and other factors. The workshop will also address current trends in climate communication, such as the use of political internet memes in online discussions on social media.

October 12, 2023 | bidt Conference Poster

From October 11 to 12, the bidt Conference 2023 took place in Munich. The conference theme was „Communicating Digitally – Communicating the Digital.“ During the conference, Simon Lübke and Johannes Fischer from the KLIMA-MEMES team presented the topic and the current status of the research project through a poster and engaged in discussions with interested participants about the project. Attendees also had the opportunity to vote for their favorite climate meme from a selection of four different memes during the conference.

September 19, 2023 | KONVENS Conference Poster

Siyao (Logan) Peng and Barbara Plank from MaiNLP and part of the KLIMA-MEMES team attended the last annual conference of computational linguistics KONVENS in Ingolstadt and presented a research poster. Their submission is titled „Adapting Named Entity Recognition to Climate Change Data“ and focuses on the method of Named Entity Recognition (NER), which the authors apply in their study to a corpus on climate change .

Short abstract: The study discusses the importance of Named Entity Recognition (NER) in Natural Language Processing (NLP) for various tasks. It also highlights the emerging role of NER in the context of Climate Change (CC) and introduces ongoing research on adapting NER to the CC domain. The study involves creating a new Climate-NER corpus, conducting experiments, and exploring domain-adaptation strategies. The research aims to enhance NER models for the CC domain, providing valuable insights into the interplay between domains, genres, and languages in model adaptation.

Peng, S. & Plank, B. (2023). Adapting Named Entity Recognition to Climate Change Data. Poster presented at CPSS@KONVENS 2023.

September 06, 2023 | Excited to see the memes – Interview about Olaf Scholz and Memes

The German chancellor, Olaf Scholz, generated excitement across social media with a photo he shared after a jogging accident. The image depicted him wearing a black eyepatch, accompanied by a caption that read, „Thanks for the well wishes. It looks worse than it is!“ He playfully added, „Excited to see the memes.“ In a recent interview conducted by the German news website DerWesten, our colleague from KLIMA-MEMES, Jörg Haßler, delves into Olaf Scholz’s social media strategy, discusses the potential impact of this post, and underscores the significance of internet memes in the realm of political communication.

The interview was conducted by Marisa Lattemann and can be accessed via the following link:

Lattemann, M. (2023, September 6). Ist das sein Merkel-Moment? Kanzler Scholz plötzlich sympathisch. DER WESTEN.

On August 23, 2023, the German newspaper Süddeutsche Zeitung (SZ) published an article on German politicians’ strategies and practices of self-presentation in social media. The article by Sonja Dawson also refers to current research results by KLIMA-MEMES colleague Jörg Haßler on the visual self-representation of parties and politicians on Instagram. The study by the LMU scholars Jörg Haßler, Anna-Katharina Wurst, and Katharina Pohl mentioned in the article was recently published in 2023 in the Journal Information, Communication & Society. The study shows that personal depictions of parties and candidates are common in Instagram campaigns while addressing policy issues in posts varies between elections. The study and the newspaper article refering to its results can be accessed via the following links:

Haßler, J.; Wurst, A.-K.; Pohl, K. (2023). Politicians over issues? Visual personlization in three Instagram election campaigns. Information, Communication & Society, 1–21.

Dawson, S. (2023, 23. August). Politiker und Fotografie: Das sind wir uns wert. Süddeutsche Zeitung. 

The Bavarian state election will be held on October 8 this year. In light of these events, the KLIMA-Memes colleague Simon Lübke was interviewed twice on the topic of visualization of political communication. on social media. The feature on self-presentation in politics explores the question whether or not it is reasonable for politicians to spend more money on professional social media performances. The video produced by Julia Brestrich was released on 08 August 2023 and can be accessed via the following link:

In a feature of the public service news magazine, ARD-Mittagsmagazin.

Brestrich, J. (2023, 8. August). Selbstinszenierung in der Politik. ARD mittagsmagazin.

In an interview with the Mittelbayerische Zeitung (MZ), he commented on the relevance and effects of election campaign posters and the strategies of German parties in the city of Regensburg. The article by Juliana Ried with the interview was published on August 3 and can be accessed via the following link:

Ried, J. (2023, 3. August). Sind diese Regensburger Wahlplakate Stimmenbringer? Mittelbayerische Zeitung.

Dr. Lea Frermann from the University of Melbourne visited Prof. Barbara Plank and the MaiNLP research lab today and presented her research on “Conflicts, Villains, Resolutions: Towards models of Narrative Media Framing”. Her talk focused on the application of NLP approaches for automatic frame prediction in the climate change discourse.

Abstract: Stories have existed as long as human societies, and are fundamental to communication, culture, and cognition. This talk looks at the interaction of narratives and media framing, i.e., the deliberate presentation of information to elicit a desired response or shift in the reader’s attitude. While rich theories of media framing have emerged from the political and communication sciences, NLP approaches to automatic frame prediction tend to oversimplify the concept. In particular, current approaches focus on overly localized lexical signals, make unwarranted independence assumptions, and ignore the broader, narrative context of news articles. This talk presents our recent work which incorporates narrative themes, roles of involved actors, and the interaction multiple frames in a news article as a step towards a computational framework of narrative framing. Quantitative evaluation and case studies on media framing of climate change reflect a benefit of the more nuanced emerging frame representations.

Bio: Dr. Lea Frermann is a lecturer (assistant professor) and DECRA fellow at the University of Melbourne. Her research combines natural language processing with the cognitive and social sciences to understand how humans learn about and represent complex information and to enable models to do the same in fair and robust ways. Recent projects include models of meaning change; of common sense knowledge in humans and language representations; and automatic story understanding in both fiction (books or movies) and the real world (as narratives in news reporting on complex issues like climate change).

Prof. Dr. Mario Haim and Dr. Simon Lübke attended a workshop on Climate Change Visuals at the Technical University of Munich today. The workshop was hosted by Stefanie Walter, PhD (TU Munich) and Prof. Dr. Mike S. Schäfer (University of Zurich) and brought together international social and computer scientists who work with computational approach to analyze climate change visuals. The participants discussed challenges and opportunities of communication climate change via images and shared experiences and insights on visualizing climate change data in social sciences research and practice.

Workshop summary: Computational methods are becoming increasingly popular among social scientists. Large amounts of data are being generated and made publicly available every day. However, it is still only a small number of researchers who have the methodological knowledge or are skilled in using computational methods which make this data so valuable. Large scale initiatives, such as quanteda have lowered the barriers to use Natural Language Processing and quantitative text analysis methods. However, these endeavours have to a very large extent been limited to textual data, following the traditional approach to content analysis that is focused on texts. Images and advances in the area of image analysis have been neglected, leaving gaps in the potential that these data sources offer. This is a big omission given that images are especially prevalent in online environments and on social media in particular. In this workshop, we want to bring together scholars working on image analysis (especially with focus on climate change) and discuss questions related to data access and analysis. Regarding the latter, we are particularly interested in exploring the challenges and opportunities that collaborations with computer scientists bring.

May 8, 2023 | Talk by Dr. Saif M. Mohammad 

Dr. Saif M. Mohammad from the National Research Council Canada (NRC) visited Prof. Barbara Plank and the MaiNLP research lab today and presented his research on “The Search for Emotions, Creativity, and Fairness in Language“. Dr. Mohammad is a leading expert in computational linguistics and his research focuses on emotions in Language. He serves as an external consultant for the KLIMA-MEMES project. 

Abstract: Emotions are central to human experience, creativity, and behavior. They are crucial for organizing meaning and reasoning about the world we live in. They are ubiquitous and everyday, yet complex and nuanced. In this talk, I will describe our work on the search for emotions in language — by humans (through data annotation projects) and by machines (in automatic emotion and sentiment analysis systems). I will outline ways in which emotions can be represented, challenges in obtaining reliable annotations, and approaches that lead to high-quality annotations and useful sentiment analysis systems. I will discuss wide-ranging applications of emotion detection in natural language processing, psychology, social sciences, digital humanities, and computational creativity. Along the way, I will discuss various ethical considerations involved in emotion recognition and sentiment analysis — the often unsaid assumptions and the real-world implications of our choices.

Bio: Dr. Saif M. Mohammad is a Senior Research Scientist at the National Research Council Canada (NRC). He received his Ph.D. in Computer Science from the University of Toronto. Before joining NRC, he was a Research Associate at the Institute of Advanced Computer Studies at the University of Maryland, College Park. His research interests are in Natural Language Processing (NLP), especially Lexical Semantics, Emotions and Language, Computational Creativity, AI Ethics, NLP for psychology, and Computational Social Science. He is currently an associate editor for Computational Linguistics, JAIR, and TACL, and Senior Area Chair for ACL Rolling Review. His word–emotion resources, such as the NRC Emotion Lexicon and VAD Lexicon, are widely used for analyzing emotions in text. His work has garnered media attention, including articles in Time, SlashDot, LiveScience, io9, The Physics arXiv Blog, PC World, and Popular Science. 

The KLIMA-MEMES project is funded by the Bavarian Research Institute for Digital Transformation.