Article Per Year (5 Year)
p-Index From 2021 - 2026
0.702
P-Index
This Author published in this journals
All Journal Journal of Business and Political Economy: Biannual Review of The Indonesian Economy Review Nusantara Journal of Artificial Intelligence and Information Systems Journal of Computing and Smart Ecosystems International Journal of Kita Kreatif
Zia Ul Rehman Zafar
Universitas Muhammadiyah Surakarta
Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Electrical & Electronics Engineering Social Sciences

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
Sentiment Analysis of Emotional Intensity as a Continuous Driver of Engagement and Algorithmic Visibility Zia Ul Rehman Zafar; Dedi Gunawan; Muhammad Saif
Nusantara Journal of Artificial Intelligence and Information Systems Vol. 2 No. 1 (2026): June
Publisher : Faculty of Engineering and Computer Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47776/nuai.v2i1.2006

Abstract

This study investigates how emotional intensity, rather than sentiment direction, shapes engagement and algorithmic visibility in digital political discourse. Using sentiment analysis, a dataset of about 15,000 posts from Twitter (X) and YouTube was collected over a 30-day period and scored with a hybrid TextBlob, VADER, and BERT pipeline. Emotional strength (the absolute sentiment value) correlated moderately with engagement (r = 0.58, p < 0.05), whereas the directional sentiment score did not (r ≈ 0.05). Emotionally intense posts attracted about 2.4 times more engagement than neutral posts, and positive posts were the most frequent (41%) while neutral posts drew the lowest mean engagement. These results indicate that engagement-based ranking amplifies emotional magnitude over neutral or analytical content, which can narrow the diversity of visible expression. The findings give platform designers and policymakers a reproducible basis for assessing how affective dynamics shape visibility in algorithmically mediated public discourse.
Informational Complexity and Market Structure: A Nonlinear Panel Analysis of Stock Returns Using Entropy and Kullback–Leibler Divergence Zia Ul Rehman Zafar; Ahad Sultan; Kashif Ali Abdul Wahid Alias; Ahmad Hassan
International Journal of Kita Kreatif Vol 3, No 2 (2026): International Journals Kita Kreatif Vol. 3 No.2 Mei 2026
Publisher : Universitas Syiah Kuala

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

This study examines how informational complexity and market structure jointly influence stock returns in an emerging market context. Using panel data from 40 publicly listed firms across eight sectors during 2005–2026, the study measures informational complexity through Shannon entropy and Kullback–Leibler (KL) divergence, while market concentration is captured using the Herfindahl–Hirschman Index (HHI). Fixed effects panel regression models are employed to evaluate the direct and interaction effects of informational and structural variables, with additional macroeconomic controls including inflation, interest rates, and exchange rates. The results show that informational complexity is positively associated with stock returns in baseline models, indicating that uncertainty and distributional irregularities contain economically relevant information. However, the significance of these effects declines after incorporating macroeconomic controls and time effects, suggesting that informational dynamics are partly shaped by broader economic conditions. The interaction between entropy and market concentration remains positive and robust, demonstrating that the effect of informational complexity depends on market structure and becomes stronger in more concentrated sectors. Exchange rates also exhibit a significant negative relationship with stock returns, highlighting the importance of macroeconomic stability in emerging markets. The study is limited by its focus on a single emerging market and the use of annual aggregated data, which may not fully capture higher-frequency dynamics. Nevertheless, the findings provide practical implications for investors and policymakers by emphasizing the conditional nature of informational effects in financial markets. The study’s originality lies in integrating information-theoretic measures and market structure within a unified empirical framework to explain nonlinear stock return dynamics.
Financial News Sentiment and Market Stability in Indonesia: A Comparative ASEAN Analysis (NLP) Zia Ul Rehman Zafar; Muhammad Saif; Mohammad Panah Alias Faraz Ahmed Ahmed; Muhammad Arsalan; Muhamma Nouman
Journal of Business and Political Economy : Biannual Review of The Indonesian Economy Vol. 8 No. 1 (2026): Journal of Business and Political Economy: Biannual Review of The Indonesian Ec
Publisher : INDEF - Institute for Development of Economics and Finance

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46851/319

Abstract

The rapid expansion of digital financial information has increased the influence of news-driven narratives on capital market behavior in emerging ASEAN economies. This study examines how financial news sentiment affects stock returns and conditional volatility in Indonesia, Malaysia, and Singapore during 2015–2024. A daily sentiment index was constructed from Reuters, Bloomberg, Factiva, and LexisNexis articles using a hybrid Natural Language Processing (NLP) approach combining lexicon-based methods and FinBERT classification. The sentiment measures were integrated into panel regression, correlated random effects (CRE), Granger causality, and GARCH(1,1) models. The results show that sentiment significantly affects both returns and volatility across ASEAN markets. A one-standard-deviation decline in sentiment increases conditional volatility by approximately 10.3% in Indonesia (p < 0.01), with weaker effects observed in Malaysia and Singapore. The return estimations indicate that a one-unit increase in sentiment raises next-day returns by approximately 0.084 percentage points. Negative sentiment generates stronger volatility responses than positive sentiment, supporting behavioral asymmetry and loss-aversion interpretations. Cross-country findings further show that sentiment sensitivity is strongest in Indonesia and weakest in Singapore, suggesting that institutional development moderates the transmission of digital information into market outcomes. The study contributes by integrating FinBERT-based sentiment analysis with comparative ASEAN financial econometrics and demonstrates the growing importance of narrative-driven risk in emerging capital markets. Keywords: Financial news sentiment; Market stability; Indonesia; ASEAN equity markets JEL Classification: G14; G15; C58
Mechanistic Interpretability of Transformer Attention Heads in Indonesian Fiscal Policy Discourse: Decoding Internal Representations and Latent Biases Zia Ul Rehman Zafar; Dedi Gunawan; Kashif Ali Abdul Wahid Alias; Ahmad Hassan; Adnan Nadeem
Journal of Computing and Smart Ecosystems Vol. 2 No. 1 (2026): J-CaSE
Publisher : S1 Teknologi Informasi, Universitas Muhammadiyah Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Transformer-based language models are increasingly used in economic text analysis, yet their internal reasoning remains insufficiently interpretable, particularly in multilingual policy contexts. This study examines the mechanistic interpretability of attention heads in analyzing Indonesian fiscal policy discourse using a bilingual dataset comprising 47 YouTube transcripts and 23,500 news articles collected between 2018 and 2025. A fine-tuned IndoBERT model is implemented for fiscal stance classification, achieving strong performance with accuracy of 0.91, precision of 0.89, recall of 0.88, and F1-score of 0.885, ensuring a reliable basis for interpretability analysis. To uncover internal model behavior, an integrated framework combining attention entropy, head importance scoring, and Bias Sensitivity Index (BSI) is applied. Results show a consistent decline in attention entropy across layers from 2.85 to 1.62, indicating a shift from distributed contextual processing to focus semantic representation. Head-level analysis reveals functional specialization, with key attention heads reaching importance scores up to 0.25. Ablation experiments further demonstrate measurable prediction shifts, reflected in elevated BSI values, confirming sensitivity to specific heads associated with fiscal risk terms. These findings indicate that transformer models encode structured economic reasoning while exhibiting latent bias, highlighting the need for interpretability-driven approaches in AI-based policy analysis.
Co-Authors Adnan Nadeem Ahad Sultan Ahmad Hassan Ahmad Hassan Dedi Gunawan Dedi Gunawan Kashif Ali Abdul Wahid Alias Kashif Ali Abdul Wahid Alias Mohammad Panah Alias Faraz Ahmed Ahmed Muhamma Nouman Muhammad Arsalan Muhammad Saif Muhammad Saif