The reduction of dissolved organic matter (DOM) in palm oil mill effluent (POME) was investigated by using electrocoagulation (EC) with aluminum electrodes. The operational parameters, including current intensities (12, 14, 16, and 31 A), applied voltages (2 V and 4 V), and contact times (30, 60, 120, and 180 min), were examined to assess their influence on DOM removal during EC treatment. UV absorbance at 260 nm (UV260), dissolved organic carbon (DOC), and fluorescent DOM were monitored to assess the changes in concentration and composition. The raw POME contained 2687 mg/L of DOC and exhibited high levels of protein-like, fulvic-like, and humic-like components. The results showed that EC at 31 A with a stirring speed of 30 rpm achieved reductions of 83% in DOC, 90% in UV260, and 67% in humic-like Peak 5 after 45 min. Additionally, the aromatic content of DOM, measured by specific UV absorbance (SUVA), decreased by 62% after 15 min of treatment at 31 A with 10 rpm stirring, indicating that DOM transformed into aliphatic compounds. The Peak 5/Peak 2 ratio showed a good correlation with SUVA, suggesting that EC effectively reduced aromatic humic-like DOM. Furthermore, X-ray fluorescence analysis of the resulting EC sludge revealed a high aluminum content, both in its elemental form (60.4%) and as Al2O3 (71.3%), suggesting potential for material recovery and reuse.
Harmful online content can negatively influence users and create social risks. This study develops a machine learning model to detect harmful website content using Naïve Bayes, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). The analysis focuses on website text extracted from the HTML Document Object Model (DOM). Four classes were used: gambling, pornography, phishing, and whitelist content. The dataset consisted of 2911 URLs collected from the UT1 Blacklist repository. Text preprocessing and TF-IDF feature extraction with unigram and bigram representations produced 71,967 tokens. Experimental results show that SVM achieved the best performance with 90.50% accuracy on 2821 active URLs. A real-time Flask-based web application was also developed to classify URLs from user input. The findings demonstrate that combining NLP and machine learning provides an effective and practical solution for harmful website content detection.
Social media platforms expose many Generation Z learners to English outside the classroom, raising the question of whether such incidental encounters are associated with measurable language ability. This study examined whether the frequency of incidental English learning (IL) through social media is associated with English proficiency among Indonesian secondary-school learners of English as a foreign language. A cross-sectional survey design was used. IL frequency was measured with a 25-item Likert-scale questionnaire capturing self-reported encounters with English on platforms such as YouTube, Instagram, and TikTok, while English ability was assessed with a 45-item test of listening, vocabulary, and grammar. Descriptive statistics, Pearson and Spearman correlations (with bootstrap confidence intervals), and linear regression were used to estimate the zero-order relationships between IL frequency and overall as well as sub-skill scores. Across all analyses, IL frequency showed no statistically significant association with total proficiency or with listening, vocabulary, and grammar sub-scores, and confidence intervals suggested that any undetected positive relationships, if present, are likely to be small. These findings, within the limits of a self-report, cross-sectional design, and a single EFL context, suggest that frequent incidental exposure via social media may not be sufficient on its own to produce measurable gains in formal proficiency. The study is preliminary and does not address causal effects, but it provides a baseline for future longitudinal and theory-rich work that considers not only how often, but also how and why adolescents engage with English on social media.
This study examines the use of intrapersonal directive speech acts as an instrument of inner speech to build psychological resilience within Japanese song lyrics. Employing a descriptive-analytical qualitative approach, this research integrates John R. Searle’s Speech Act Theory and Lev Vygotsky’s Inner Speech Theory to analyze four selected songs: Homura and Gurenge by LiSA, alongside Unravel and Signal by TK from Ling tosite sigure. The analysis of 37 utterance data reveals that the dynamics of this inner communication are predominantly characterized by the Self-Command function (45.9%), which is oriented towards behavioral activation and reinforcing determination, as well as the Self-Advice function (40.5%), which serves the purpose of cognitive restructuring. Furthermore, the study identifies Self-Prohibition speech acts (8.1%), acting as an ego defense mechanism during acute crises, and Self-Request (5.4%), which functions as a preliminary step for emotional validation. In conclusion, directive utterances in Japanese song lyrics transcend mere passive emotional expression; they effectively transform into a hierarchical and synergistic mechanism of intrapersonal communication (self-talk) utilized to convert trauma or sorrow into an internal motivational drive to overcome adversity.
Persistent workload pressure in public healthcare institutions raises concerns about declining job satisfaction, particularly when excessive job demands disrupt the balance between professional and personal roles. This study investigates the direct and indirect effects of work overload on job satisfaction through the mediating mechanisms of work-life conflict among nurses at RSUD Dr. Muhammad Zein Painan, a regional public hospital in West Sumatra, Indonesia. Grounded in the Job Demands-Resources theory, work-life conflict is conceptualized as a reflective-reflective second-order construct comprising Work Interference with Life (WIL) and Life Interference with Work (LIW). A quantitative explanatory design was employed using purposive sampling, involving 158 nurses assigned to inpatient wards and the emergency department. Data were collected through a five-point Likert scale questionnaire and analyzed using Structural Equation Modeling with the Partial Least Squares (SEM-PLS) approach. The findings indicate that work overload exerts a negative and significant direct effect on job satisfaction and positively influences both WIL and LIW, suggesting intensified bidirectional role interference. However, only LIW demonstrates a negative and significant effect on job satisfaction, whereas WIL does not show a significant direct relationship. Mediation analysis reveals that LIW provides complementary mediation in the relationship between work overload and job satisfaction, while WIL exhibits direct-only non-mediation. Higher-order construct evaluation confirms that both dimensions significantly form work-life conflict, with LIW showing a slightly stronger contribution. These results demonstrate an asymmetric transmission mechanism in which reverse role interference plays a more critical role in translating workload pressure into reduced job satisfaction. The study refines demand-based explanations of employee attitudes in public healthcare contexts and emphasizes the strategic importance of workload management and institutional support systems in sustaining nurses’ job satisfaction.
