NLP Workflows for Computational Social Science: Understanding Triggers of State-Led Mass Killings

Burley, Timothy; Humble, Lorissa; Sleeper, Charles; Sticha, Abigail; Chesler, Angela; Regan, Patrick M.; Verdeja, Ernesto

doi:10.1145/3311790.3397343

Cited by 3 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These dictionary files are unique to a given trigger and follow the CAMEO standard for political event extraction (Gerner et al, 2002). We employed the trigger coding definitions from Burley et al (2020) which included the specific key words for coups. During the dictionary creation process, we found that creating new dictionaries for each refinement of a search is labor intensive and risks added bias.…”

Section: Event Coding With Petrarchmentioning

confidence: 99%

“…We switched to assassination events to create our SPE dataset because we saw a lack of assassination datasets in literature (Section 2.1) and assassinations are the most clearly defined trigger 5 for the triggers laid out in (Burley et al, 2020) with one keyword ('assassination'). We leveraged our flushed out machine learning pipeline, along with existing assassination datasets, KGs, and NER to enhance our new assassinations dataset (Fig 8 ).…”

Section: Hybrid Knowledge Engineering Tomentioning

confidence: 99%

“…However, the timing of mass killing onset is less understood. Burley et al (2020) identifies nine potential triggering events for state mass killings, such as coups and assassinations, but before socio-political researchers can conduct systematic analysis to examine whether, and if so when, certain patterns of trigger-type events actually increase the probability of mass killings, it is necessary for political researchers to obtain political event datasets for each of these potential triggering events.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Hybrid Knowledge Engineering to Build a Global Assassination Dataset

Sticha

Broussard

Havenaar

et al. 2022

2022 IEEE Fifth International Conference on Artificial Intelligence and Knowledge Engineering (AIKE)

View full text Add to dashboard Cite

Social and political researchers require robust event datasets to conduct data-driven analysis, an example being the need for trigger event datasets to analyze under what conditions and in what patterns certain trigger-type events increase the probability of mass killings. Fortunately, NLP and ML can be leveraged to create these robust datasets. In this paper we (i) outline a robust ML framework that prioritizes understandability through visualizations and generalizability through the ability to implement different ML algorithms, (ii) perform a comparative analysis of these ML tools within the framework for the coup trigger, (iii) leverage our ML framework along with a unique combination of NLP tools, such as NER and knowledge graphs, to produce a dataset for the the assassination trigger, and (iv) make this comprehensive, consolidated, and cohesive assassination dataset publicly available to provide temporal data for understanding political violence as well as training data for further sociopolitical research.

show abstract

Section: Event Coding With Petrarchmentioning

confidence: 99%

Section: Hybrid Knowledge Engineering Tomentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hybrid Knowledge Engineering to Build a Global Assassination Dataset

Sticha

Broussard

Havenaar

et al. 2022

2022 IEEE Fifth International Conference on Artificial Intelligence and Knowledge Engineering (AIKE)

View full text Add to dashboard Cite

show abstract

“…Milli and Bamman (2016) present results in a comprehensive journal article demonstrating gender differences in political discussions on Twitter, using computational methods. Other papers that rely heavily on scalable data analysis and computational methods, but with aims that are similar to more traditional social science research, include work by Rodriguez and Storer (2020); Burley et al (2020); Mason et al (2014); Keuschnigg et al (2018). These works cover a broad range of subjects in sociology, from analysis of social media (Rodriguez and Storer, 2020) and social computing (Mason et al, 2014), to a study of state-led mass killings (Burley et al, 2020), and analytical sociology (Keuschnigg et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

Quantifying Gender Disparity in Pre-Modern English Literature using Natural Language Processing

Kejriwal¹,

Nagaraj²

2023

Journal of Data Science

View full text Add to dashboard Cite

Research has continued to shed light on the extent and significance of gender disparity in social, cultural and economic spheres. More recently, computational tools from the data science and Natural Language Processing (NLP) communities have been proposed for measuring such disparity at scale using empirically rigorous methodologies. In this article, we contribute to this line of research by studying gender disparity in 2,443 copyright-expired literary texts published in the pre-modern period, defined in this work as the period ranging from the beginning of the nineteenth through the early twentieth century. Using a replicable data science methodology relying on publicly available and established NLP components, we extract three different gendered character prevalence measures within these texts. We use an extensive set of statistical tests to robustly demonstrate a significant disparity between the prevalence of female characters and male characters in pre-modern literature. We also show that the proportion of female characters in literary texts significantly increases in female-authored texts compared to the same proportion in male-authored texts. However, regression-based analysis shows that, over the 120 year period covered by the corpus, female character prevalence does not change significantly over time, and remains below the parity level of 50%, regardless of the gender of the author. Qualitative analyses further show that descriptions associated with female characters across the corpus are markedly different (and stereotypical) from the descriptions associated with male characters.

show abstract

Applying Machine Learning to Online Data?

Bayram

2022

Opportunities and Challenges for Computational Social Science Methods

View full text Add to dashboard Cite

The immense impact of social media on contemporary cultural evolution is undeniable, consequently declaring them an essential data source for computational social science studies. Alongside the advancements in natural language processing and machine learning disciplines, computational social science researchers continuously adapt new techniques to the data collected from social media. Although these developments are imperative for studying the sociological transformations in many communities, there are some inconspicuous problems on the horizon. This chapter addresses issues that may arise from the use of social media data, like biased models. It also discusses various obstacles associated with machine learning methods while also providing possible solutions and recommendations to overcome these struggles from an interdisciplinary perspective. In the long term, this chapter will guide computational social science researchers in their future studies, from things to be aware of with data collection to assembling an accurate experimental design.

show abstract

NLP Workflows for Computational Social Science: Understanding Triggers of State-Led Mass Killings

Cited by 3 publications

References 28 publications

Hybrid Knowledge Engineering to Build a Global Assassination Dataset

Hybrid Knowledge Engineering to Build a Global Assassination Dataset

Quantifying Gender Disparity in Pre-Modern English Literature using Natural Language Processing

Applying Machine Learning to Online Data?

Contact Info

Product

Resources

About