Beyond Polarity: A Python-Based Corpus Framework for Examining Foreign Media Coverage of Africa

DOI: https://doi.org/10.62517/jbdc.202601325

Author(s)

Xingchi Ma

Affiliation(s)

Xi'an Fanyi University, Xi'an, Shaanxi, China

Abstract

This paper develops a Python-based corpus framework for examining sentiment and representation in foreign media coverage of Africa. Rather than equating negative vocabulary with bias, the framework distinguishes event-related negativity, editorial evaluation, topic selection, source attribution, and narrative agency. A multi-outlet corpus is organized by headline, standfirst, article body, quotation, country, topic, and publication. VADER and TextBlob provide transparent baseline scores, while human annotation addresses irony, negation, attribution, and domain-specific language. Topic and outlet comparisons are controlled for article volume and issue mix, and concordance reading links numerical patterns to their textual contexts. The study proposes a layered route from lexical polarity to claims about framing: automated results identify clusters and anomalies; contextual coding explains who evaluates whom; comparative analysis tests whether patterns are specific to Africa or common to international reporting. The framework offers a reproducible foundation for studying foreign media narratives without treating Africa or foreign media as homogeneous categories.

Keywords

Africa; Foreign Media; Sentiment Analysis; Corpus Linguistics; Media Framing; Python

References

[1] Scott M. The myth of representations of Africa: A comprehensive scoping review of the literature[J]. Journalism Studies, 2017, 18(2):191210. [2] Nothias T. How Western journalists actually write about Africa: Reassessing the myth of representations of Africa [J]. Journalism Studies, 2018, 19 (8):11381159. [3] Galtung J, RUGE M H. The structure of foreign news[J]. Journal of Peace Research, 1965, 2(1):6490. [4] Harcup T, O'Neill D. What is news? News values revisited (again)[J]. Journalism Studies, 2017, 18 (12):14701488. [5] Bunce M, Franks S, Paterson C, eds. Africa's Media Image in the 21st Century [M]. London: Routledge, 2017. [6] Taboada M, Brooke J, Tofiloski M, voll K, Stede M. Lexiconbased methods for sentiment analysis[J]. Computational Linguistics, 2011, 37(2):267307. [7] Hutto C J, Gilbert E. VADER: A parsimonious rulebased model for sentiment analysis of social media text[C]//Proceedings of the International AAAI Conference on Web and Social Media. 2014:216225. [8] Young L, Soroka S. Affective news: The automated coding of sentiment in political texts[J]. Political Communication, 2012, 29(2):205231. [9] Entman R M. Framing: Toward clarification of a fractured paradigm[J]. Journal of Communication, 1993, 43(4):5158. [10] Yang J, Cui H, Zhang Y, et al.From “Angels” to “Role Models of the Times”: A Discourse-Historical Analysis of Ningxia’s Media Coverage of Medical Aid to Benin (2006–2024)[J].Scientific and Social Research,2026,8(3):150-157. [11] Hong A, Hu D. Analyzing Foreign Media Coverage of China During the 2022 Beijing Winter Olympics Opening and Closing Ceremonies: A Corpus-Assisted Critical Discourse Analysis[J].Journalism and Media,2025,6(3):145-145. [12] KumarP. Analysing foreign media coverage of India's Chandrayan‐3 mission[J].Asian Politics & Policy,2024,16(2):294-297. [13] Li Y ,Ruan X. Exploring the Role of Media Framing in the Self-construction of China’s National Image: An Analysis of News Coverage About Foreign Language Education in China Daily[J].The Educational Review, USA,2025,9(1):103-111..