The Sentiment is in the Details

Vries, Erik de

doi:10.5117/ccr2022.2.003.vrie

Cited by 4 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, Palmer, Roeder, and Muntermann (2021) build a domain-specific dictionary by assigning word polarity through a linear regression linking words and stock returns. De Vries (2022), following an approach introduced by Rheault et al (2016), uses a seed dictionary and a word embedding model to automatically identify additional words and their corresponding tonality. They are able to show considerable improvements of their approach compared with other dictionaries.…”

Section: Introductionmentioning

confidence: 99%

Introducing an Interpretable Deep Learning Approach to Domain-Specific Dictionary Creation: A Use Case for Conflict Prediction

Häffner¹,

Hofer²,

Nagl

et al. 2023

Polit. Anal.

View full text Add to dashboard Cite

Recent advancements in natural language processing (NLP) methods have significantly improved their performance. However, more complex NLP models are more difficult to interpret and computationally expensive. Therefore, we propose an approach to dictionary creation that carefully balances the trade-off between complexity and interpretability. This approach combines a deep neural network architecture with techniques to improve model explainability to automatically build a domain-specific dictionary. As an illustrative use case of our approach, we create an objective dictionary that can infer conflict intensity from text data. We train the neural networks on a corpus of conflict reports and match them with conflict event data. This corpus consists of over 14,000 expert-written International Crisis Group (ICG) CrisisWatch reports between 2003 and 2021. Sensitivity analysis is used to extract the weighted words from the neural network to build the dictionary. In order to evaluate our approach, we compare our results to state-of-the-art deep learning language models, text-scaling methods, as well as standard, nonspecialized, and conflict event dictionary approaches. We are able to show that our approach outperforms other approaches while retaining interpretability.

show abstract

Section: Introductionmentioning

confidence: 99%

Introducing an Interpretable Deep Learning Approach to Domain-Specific Dictionary Creation: A Use Case for Conflict Prediction

Häffner¹,

Hofer²,

Nagl

et al. 2023

Polit. Anal.

View full text Add to dashboard Cite

show abstract

Concept and challenges of a social media observatory as a DIY research infrastructure

Wiedemann,

Münch,

Rau

et al. 2023

Publizistik

View full text Add to dashboard Cite

This article describes the basic concept, ethical and legal considerations, technical implementation as well as resulting tools and data collections of the Social Media Observatory (SMO). Since 2020, the SMO is developed as an open science research infrastructure within the Research Institute Social Cohesion (RISC) in Germany. It focuses on (the support of) long-term monitoring of public communication on selected platforms and online news media to answer social science-related research questions. Based on systematically compiled lists of public speaker categories, such as parliamentarians or media organizations, it collects statistics as well as content data to study the German social media discourse in comparison to mass media. Aggregated results are published via interactive dashboards. Raw data is published as ID lists for reproduction or shared with researchers upon request. Following a do-it-yourself approach to infrastructure, the SMO further provides various tools, curated datasets, and documented workflows, for instance, to run thematic ad-hoc data collections. As a main feature, it maintains a curated knowledge base in wiki format to enable other researchers to perform systematic social media observations on their own.

show abstract

Still going strong? The role of traditional media in the 2021 Dutch parliamentary elections

2022

View full text Add to dashboard Cite

Previous research has demonstrated that both visibility of parties, party leaders, candidates, and topics, and the sentiment of this coverage can affect people’s decision in the ballot box. Most of this research was, however, done in the period before social media gained importance which has drastically changed the media consumption of citizens. The main aim of this paper is to investigate whether, and if so to what extent, traditional media use during the 2021 Dutch parliamentary elections has (still) affected vote choice in this era of social media. To study this, two-wave panel survey data from the Dutch Parliamentary Election Survey (DPES) are combined with an automated content analysis of newspaper articles ( N = 35,511). We created respondent-specific content variables to conduct a linkage analysis. Our analysis, relying on a pooled analysis of respondent–party combinations ( N = 54,162), demonstrates that political parties profit electorally from being visible in both newspapers and online outlets. This is in particular true for parties that are not part of parliament yet, thus increasing the further fragmentation and division in Dutch politics. Contrary to the expectations, sentiment in online media has a negative effect, with negative coverage increasing electoral support. Supplementary Information The online version contains supplementary material available at 10.1057/s41269-022-00270-7.

show abstract

The Sentiment is in the Details

Cited by 4 publications

References 0 publications

Introducing an Interpretable Deep Learning Approach to Domain-Specific Dictionary Creation: A Use Case for Conflict Prediction

Introducing an Interpretable Deep Learning Approach to Domain-Specific Dictionary Creation: A Use Case for Conflict Prediction

Concept and challenges of a social media observatory as a DIY research infrastructure

Still going strong? The role of traditional media in the 2021 Dutch parliamentary elections

Contact Info

Product

Resources

About