Analyzing Online Political Advertisements

Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, pages 3669–3680 August 1–6, 2021. ©2021 Association for Computational Linguistics 3669 Analyzing Online Political Advertisements Danae S´anchez Villegasα Saeid Mokaramβ Nikolaos Aletrasα α Computer Science Department, University of Sheffield, UK β Emotech {dsanchezvillegas1, n.aletras}@sheffield.ac.uk saeid.mokaram@gmail.com Abstract Online political advertising is a central aspect of modern election campaigning for influenc- ing public opinion. Computational analysis of political ads is of utmost importance in polit- ical science to understand the characteristics of digital campaigning. It is also important in computational linguistics to study features of political discourse and communication on a large scale. In this work, we present the first computational study on online political ads with the aim to (1) infer the political ideol- ogy of an ad sponsor; and (2) identify whether the sponsor is an official political party or a third-party organization. We develop two new large datasets for the two tasks consisting of ads from the U.S.. Evaluation results show that our approach that combines textual and vi- sual information from pre-trained neural mod- els outperforms a state-of-the-art method for generic commercial ad classification. Finally, we provide an in-depth analysis of the limita- tions of our best-performing models and lin- guistic analysis to study the characteristics of political ads discourse.1 1 Introduction Online advertising is an integral part of modern digital election campaigning (Fulgoni et al., 2016; Fowler et al., 2020a). The increased spending on online political ads (e.g. the 2020 U.S. election campaign spending hit an all-time record2) poses a significant challenge to the democratic oversight of digital campaigning,3 with serious implications 1Data is available here: https://archive.org/de tails/pol ads 2https://www.cnbc.com/2020/10/01/elec tion-2020-campaign-spending-set-to-hit -record-11-billion.html 3https://www.electoral-reform.org.uk/ latest-news-and-research/publications/de mocracy-in-the-dark-digital-campaigning- in-the-2019-general-election-and-beyond/ about transparency and accountability, for example how voters are targeted and by whom (Kriess and Barrett, 2020). Political advertising is defined as ‘any controlled message communicated through any channel de- signed to promote the political interests of indi- viduals, parties, groups, government, or other or- ganizations’ (Kaid and Holtz-Bacha, 2006). It is guided by ideology and morals (Scammell and Langer, 2006; Kumar and Pathak, 2012), and often expresses more negativity (Haselmayer, 2019; Iyen- gar and Prior, 1999; Lau et al., 1999) compared to the aesthetic nature of commercial advertising. Ta- ble 1 shows examples of online political ads across different political parties and sponsor types. While the closely related online commercial ad- vertising domain has recently been explored in nat- ural language processing (NLP) for predicting the category (e.g. politics, cars, electronics) and sen- timent of an ad (Hussain et al., 2017; Kalra et al., 2020), online political advertising has yet to be explored. Large-scale studies of online political advertising have so far focused on understanding targeting strategies rather than developing predic- tive models for analyzing its content (Edelson et al., 2019; Medina Serrano et al., 2020). Automatically analyzing political ads is impor- tant in political science for researching the char- acteristics of online campaigns (e.g. voter tar- geting, sponsors, non-party campaigns, privacy, and misinformation) on a large scale (Scammell and Langer, …

Content truncated to 500 words