VOICES OF DISPLACED: A VADER-BASED SENTIMENT ANALYSIS OF AFGHAN REPATRIATES' LONG-TERM LIVED EXPERIENCES IN PAKISTAN

Abstract

Afghanistan's protracted displacement crisis has produced millions of long-term refugees in Pakistan, many of whom now face forced repatriation under the government's Illegal Foreigners' Repatriation Plan announced in October 2023. While existing scholarship has examined the geopolitical and legal dimensions of this displacement, the emotional and linguistic experiences of refugees confronting deportation remain underexplored through computational methods. This study addresses this gap by applying VADER-based sentiment analysis in the Orange data mining software to 20 semi-structured interview transcripts from male Afghan refugees aged 50 and above, residing at Chakdara Refugee Camp, District Dir Lower, Khyber Pakhtunkhwa, Pakistan. Interviews were conducted in Pashto from December 2025 to January 2026 and analysed through word clouds, heat maps, box plots, and sentiment distribution curves. The findings reveal coexisting expressions of joy rooted in economic contribution and generational belonging, alongside pervasive fear driven by institutional violence, legal marginalisation, and the threat of return to Taliban-controlled Afghanistan. The study recommends that policymakers incorporate refugee voices and emotional evidence into deliberations on deportation policy.