Hasil untuk "Islam. Bahai Faith. Theosophy, etc."

Susetyo Rohadi, Widi Hartono, Dewi Handayani et al.

To develop effective techniques for selecting subcontractors, this research set out to determine which factors should take precedence and which alternatives may be useful in terms of BCR, benefits, costs, and risks. Based on the results Project managers, control managers, procurement bureau managers, production managers, finance bureau managers, and employees/staff directly related to procurement activities (i.e., procurement staff) were randomly selected to participate in in-depth interviews and questionnaires to gather data for the study of subcontractor selection criteria in construction services. Analytical Network Process (ANP) using BCR network structure methodology is used to analyze the data in this study. Benefit, Cost, and Risk (BCR) study of the subcontractor selection process yielded the following results: control criteria. As a result, the decision-making process for choosing subcontractors shifts its focus to other technical competency factors, which might provide advantages.

Syamsul Aripin, Nana Meily Nurdiansyah, Armai Arief et al.

Purpose – This study aims to gain an in-depth understanding of the effectiveness and reformulation of Islamic Religious Education (IRE) in the modern era. Design/methods/approach – The research method used is a literature study with a descriptive qualitative approach. The data sources and data collection techniques used were obtained from books, journals, or scientific articles, and the analysis technique used was content analysis. Findings – The results of the study show that, first, changes to Islamic Religious Education in the modern era are very complex, because the elements of modernity continue to undergo changes in terms of teaching, material, methods, and approaches. The effectiveness of Islamic Religious Education (IRE) in the modern era requires a reformulation of learning that is contextual and adaptive. Therefore, the differences that arise due to the development of science and technology need to be seen as a gift that enriches comprehensive and holistic understanding; second, learning is directed at solving the problems of modernity faced by Muslims through scientific, social, and religious approaches; third, the delivery of Islamic knowledge is not dogmatic, but emphasizes historical analysis so that students are able to understand the dynamics of the development of Islamic teachings; fourth, text-based learning methods must be balanced with critical analysis of social realities to make them more applicable; fifth, strengthening the spiritual dimension through Sufism is a fundamental requirement in shaping religious character; sixth, the effectiveness of Islamic Education is not only measured by individual piety, but also by its contribution to building social piety.    Research implications/limitations – Data sources are limited to online and offline scientific literature based on literature reviews as the primary source. Future researchers can use more in-depth techniques and methods such as research and development or case studies. Originality/value – The results of this study provide knowledge and implications for IRE teaching models and their reformulation to contribute to the challenges and progress of the times, so that synergy between various transformations can be achieved.

Yoshiyasu Ozeki

The notion of a Kummer-faithful field, defined by Mochizuki, is expected as one of suitable base fields for anabelian geometry. In this paper, we study Kummer-faithfulness for algebraic extension fields of $p$-adic fields. We show that Kummer-faithfulness for such fields are deeply related with various finiteness properties on torsion points of (semi-)abelian varieties. For example, a Galois extension $K$ of a $p$-adic field is Kummer-faithful with finite residue field if and only if, for any finite extension $L$ of $K$ and any abelian variety over $L$,its $L$-rational torsion subgroup is finite. In addition, we study Kummer-faithfulness for Lubin-Tate extension fields.

Junxian Li, Xinyue Xu, Sai Ma et al.

Multimodal Large Language Models (MLLMs) frequently suffer from unfaithfulness, generating reasoning chains that drift from visual evidence or contradict final predictions. We propose Faithful-First Reasoning, Planning, and Acting (RPA) framework in which FaithEvi provides step-wise and chain-level supervision by evaluating the faithfulness of intermediate reasoning, and FaithAct uses these signals to plan and execute faithfulness-aware actions during inference. Experiments across multiple multimodal reasoning benchmarks show that faithful-first RPA improves perceptual faithfulness by up to 24% over prompt-based and tool-augmented reasoning frameworks, without degrading task accuracy. Our analysis shows that treating faithfulness as a guiding principle perceptually faithful reasoning trajectories and mitigates hallucination behavior. This work thereby establishes a unified framework for both evaluating and enforcing faithfulness in multimodal reasoning. Code will be released upon acceptance.

Farah Shahin

This study explores the dynamic and diverse Islamic feminist movement, which aims to promote justice and gender equality by challenging patriarchal interpretations of Islamictexts and customs. Through the work of scholars such as Fatima Mernissi, this research highlights how Islamic feminism seeks to reinterpret Islamic theology and practice to uphold the rights and dignity of women. Using exploratory and historical qualitative methods, this research evaluates the significant contribution of Islamic feminists in challenging existing discourse and improving the status of women. This study examines the contributions of Islamic feminists in Morocco and the Arab world, focusing on their efforts to promote gender equality within an Islamic framework. by investigating Fatima Mernissi's work which emphasizes the need to re-read Islam's sacred texts and expose false and misogynistic hadith. Apart from that, the influence of Kecia Ali's thoughts in "Sexual Ethics and Islam" as well as the critical views of Asma Barlas, Leila Ahmed, and Amina Wadud regarding misinterpretations of the Koran enrich this research so that it provides a comprehensive view. She argues that Fatima Mernissi's works, by offering alternative readings of Islamic discourse on women, have advanced feminist debate and practice in the Arab world, enriching understanding of faith, gender, identity and culture in Islam.

Shella Septiawati, Maulita Lutfiani

This study aims to investigate the level of emotional quotient (EQ) and its impact on the performance of employees in the Student Sports Education and Training Technical Implementation Unit for Students and Students of Bogor District. The research method employed surveys and data analysis using SPSS. The results indicate that EQ significantly influences the performance of employees in the unit, accounting for 63.8%, while other factors contribute to the remainder. Further analysis using SPSS reveals that emotional intelligence factors, such as emotional stability, emotional strength, and emotional satisfaction, have significant values ​​confirming their impact on performance. These findings provide valuable insights for human resource management in understanding the importance of EQ in enhancing the productivity and quality of employee performance within the organizational environment.

بشرى إسماعيل أحمد أرنوط

Structural equation modeling using the Partial Least Square-Structural Equation Modeling algorithm (PLS-SEM) is a statistical technique that has gained attention recently due to its flexibility and predictive power. However, there is a scarcity of guidance on applying this technique in the field of social sciences, especially education and psychology research. Therefore, this study aimed to apply the PLS-SEM algorithm as an advanced approach to structural equation modeling, highlight the justifications for its use and compare it to structural modeling using the variance method. PLS-SEM was applied using the SmartPLS program to a proposed structural model of the causal relationships between artificial intelligence use, therapeutic alliance, and job engagement. The descriptive approach was applied. The study sample consisted of (127) mental health service providers in the Kingdom of Saudi Arabia, including 58 males and 69 females aged 25-50 years. (36.32±6.43), the artificial intelligence questionnaire, therapeutic alliance scale, and job engagement scale were applied to them, all prepared by the researcher. The results found there were median levels of artificial intelligence use, therapeutic alliance, and job engagement, and also showed that the proposed structural model of artificial intelligence use and the therapeutic alliance has a good ability to predict job engagement and explain the interrelationships between them compared to the indicators model and the linear model. The results also revealed a strong overall positive effect statistically significant (p< 0.05) for the variable of attitude towards using artificial intelligence in the therapeutic alliance (0.941) and job engagement (0.930), and a moderate overall positive effect statistically significant (p< 0.05) for the therapeutic alliance in job engagement (0.694). These findings indicate the importance of integrating training of mental health providers on the skills of using artificial intelligence and generative artificial intelligence techniques into professional practice to develop their ability to build a therapeutic alliance with beneficiaries and enhance their sense of the importance of the profession and their well-being in it.

جمال سلیمانی, سید محمدرضا سید نورانی

مقدمه: در شرایط کنونی اقتصاد جهانی، بانک‌ها به‌عنوان یکی از پایدارترین نهادهای اقتصادی و یکی از عوامل اصلی رشد اقتصادی شناخته می‌شوند. این پژوهش با هدف بررسی تأثیر سلامت مالی بانک‌ها، رقابت در صنعت بانکی و درآمد سرانه بر انعطاف‌پذیری مالی بانک‌های پذیرفته‌‌شده در بورس اوراق بهادار تهران انجام شده است. انعطاف‌پذیری مالی به توانایی یک بانک در مدیریت ساختار مالی خود و جذب منابع مالی مورد نیاز در طول فعالیت‌های تجاری و در مواجهه با تغییرات اقتصادی اطلاق می‌شود. دوره زمانی مورد مطالعه، یعنی از سال 1390 تا 1400، شاهد تحولات متعدد اقتصادی و سیاسی در ایران ازجمله تحریم‌ها، نوسانات شدید قیمت نفت و تغییرات سیاست‌های اقتصادی داخلی بوده است. این عوامل بر محیط رقابتی و عملکرد بخش بانکی کشور تأثیر قابل‌توجهی گذاشته‌اند.هدف اصلی این پژوهش، بررسی تأثیر سلامت مالی بانک‌ها و شدت رقابت در صنعت بانکی بر انعطاف‌پذیری مالی بانک‌های ایرانی است. افزون‌براین، پژوهش حاضر به بررسی نقش درآمد سرانه به‌عنوان یک عامل کلان اقتصادی نیز می‌پردازد که بر استقلال مالی بانک‌ها تأثیرگذار است. با شناسایی این عوامل، پژوهش تلاش می‌کند تا پیشنهادات کاربردی و راهبردی را برای سیاست‌گذاران و مدیران بانکی درجهت تقویت ثبات مالی و اتخاذ راهبرد‌های رقابتی مناسب در صنعت بانکی ایران ارائه نماید. انتظار می‌رود نتایج این پژوهش به تدوین سیاست‌های پولی و بانکی صحیح و کارآمدی منجر شود که درنهایت به رشد و پایداری اقتصاد ایران و نظام بانکی آن کمک شایانی نماید.روش تحقیق: در این پژوهش، از روش تعمیم‌یافته حداقل مربعات (GMM) در چارچوب داده‌های تابلویی پویا برای بررسی انعطاف‌پذیری مالی بانک‌های پذیرفته‌‌شده در بورس اوراق بهادار تهران در بازه زمانی 1390 تا 1400 استفاده شده است. برآوردگر GMM به‌دلیل توانایی آن در حل مسائل درون‌زایی و کنترل هتروژنیتی نادیده گرفته‌شده میان بانک‌ها، که دو چالش رایج در تحلیل داده‌های تابلویی هستند، انتخاب شده است. در ادامه، مشخصات مدل تجربی ارائه شده است:INCLUSIONit = 𝞫0 + 𝞫1 INCLUSIONit-1 + 𝞫2 DEPTHit + 𝞫3 GDPit + 𝞫4 CONCENTRATIONit + 𝞮tدر این رگرسیون متغیرهای تحقیق به‌صورت زیر تعریف می‌شوند::متغیر وابستهINCLUSIONit: شاخص انعطاف‌پذیری مالی در شرکتi  در سال tام می‌باشد.INCLUSIONit-1: شاخص انعطاف‌پذیری مالی در شرکتi  در سال t-1ام می‌باشد.DEPTHit: شاخص قدرت بانکی در شرکتi  در سال tام می‌باشد.GDPit: درآمد سرانه در شرکتi  در سال tام می‌باشد.CONCENTRATIONit: شاخص تمرکز بانکی (شاخصی برای نشان دادن رقابت بانکی) در شرکتi  در سال tام می‌باشد.داده‌های مورد استفاده در این پژوهش از صورت‌های مالی بانک‌های پذیرفته‌شده در بورس اوراق بهادار تهران و شاخص‌های کلان اقتصادی از منابع آماری ایران گردآوری شده‌اند. آزمون‌های ریشه واحد ایستایی سری‌های زمانی را تأیید کرده و آزمون‌های هم‌ترازی پنلی وجود رابطه بلندمدت بین متغیرها را تأیید نموده‌اند تا از صحت داده‌ها اطمینان حاصل شود. آزمون سارگان برای بررسی اعتبار ابزارهای مورد استفاده در روش GMM و آزمون آرلانو-باند برای بررسی خودهمبستگی سریالی خطاها انجام شده‌اند. این مراحل روش‌شناختی دقیق، اعتبار و استحکام یافته‌های پژوهش را تضمین می‌کنند.نتایج: در این مطالعه، تحلیل تجربی با استفاده از روش تعمیم‌یافته حداقل مربعات برای بررسی انعطاف‌پذیری مالی بانک‌های پذیرفته‌شده در بورس اوراق بهادار تهران طی دوره 1390 تا 1400 انجام شد. یافته‌ها نشان می‌دهد که رابطه معنا‌داری بین انعطاف‌پذیری مالی بانک‌ها و قدرت بانکی وجود دارد. این نتیجه نشان می‌دهد که بانک‌های قوی‌تر، به‌ویژه آنهایی که قدرت بیشتری در تعیین عرضه پول و تخصیص اعتبار دارند، از سیستم‌های مالی انعطاف‌پذیرتری برخوردار هستند.شاخص تمرکز بانکی نیز تأثیر مثبت و معناداری بر انعطاف‌پذیری مالی دارد. این نتیجه نیز از نظریه رقابت حمایت می‌کند که بیان می‌کند رقابت باعث پایداری بیشتر سیستم مالی می‌شود؛. زیرا رقابت بانک‌ها را مجبور می‌کند تا منابع خود را سازمان‌دهی کنند و اطمینان حاصل کنند که این اتفاق می‌افتد که تأثیر مثبتی بر نقدینگی و دسترسی به منابع مالی دارد.همچنین، درآمد سرانه نیز به‌عنوان یک عامل کلان اقتصادی مهم، تأثیر مثبت و معنا‌داری بر انعطاف‌پذیری مالی دارد. این نتیجه نشان می‌دهد که رشد اقتصادی و افزایش درآمد سرانه می‌تواند به تقویت انعطاف‌پذیری مالی بانک‌ها کمک کند. افزون‌براین، نتایج آزمون‌های تشخیصی، ازجمله آزمون سارگان و آزمون آرلانو-باند، نشان می‌دهد که مدل مشخص‌شده به‌خوبی برازش داده‌ها را انجام می‌دهد و تخمین‌های حاصل از مدل معتبر هستند.جدول 1: خلاصه نتایج برآورد  GMMمتغیرضریبخطای استانداردآماره tمقدار pانعطاف‌پذیری مالی با تأخیر0.2065430.0830882.4858330.0147قدرت بانکی0.7346800.3673701.9998210.0481درآمد سرانه0.4000890.1909842.0948880.0391تمرکز بانکی0.2522200.0836533.0150950.0032آماره J (آزمون سارگان)  8.1161510.322463نکته: معنا‌دار در سطح 5 درصد.به‌طورکلی این یافته‌ها تأثیر مثبت قدرت بانک‌ها، محیط‌های رقابتی و رونق اقتصادی بر انعطاف‌پذیری مالی بانک‌های منتخب را تأیید کرده و پیامدهای ارزشمندی را برای سیاست‌گذاران و فعّالان بانکی فراهم می‌کند.بحث و نتیجه‌گیری: نتایج این مطالعه نشان می‌دهد که عوامل موردبررسی، ازجمله قدرت بانک‌، رقابت بانکی و درآمد سرانه، تأثیر مثبتی بر انعطاف‌پذیری مالی بانک‌های پذیرفته‌شده در بورس اوراق بهادار تهران در دوره 1390 تا 1400 داشته‌اند. این امر، این دیدگاه را تقویت می‌کند که نهادهای مالی قوی در تقویت مدیریت عرضه پول و سیستم‌های تخصیص اعتبار برای بهبود انعطاف‌پذیری شرکت‌ها در طول چرخه‌های اقتصادی بسیار مهم هستند. همچنین، رابطه مثبت بین رقابت بانکی نشان می‌دهد که رقابت ازطریق تشویق بانک‌ها به مدیریت کارآمد منابع، نوآوری بیشتر و حفظ نقدینگی کافی، که برای پایداری سیستم مالی مهم است، عملکرد بانک‌ها را بهبود می‌بخشد.همبستگی مثبت بین درآمد سرانه و انعطاف‌پذیری مالی نشان می‌دهد که شرایط کلان اقتصادی بر توانایی بانک‌ها در دسترسی به منابع مالی و مقابله با ریسک‌ها تأثیر می‌گذارد. یافته‌های این مطالعه با مطالعات قبلی که بر محیط رقابتی و محیط کلان اقتصادی به‌عنوان عوامل تعیین‌کننده عملکرد و انعطاف‌پذیری مالی تأکید کرده‌اند، همسو است.براساس این مطالعه، می‌توان پیامدهای سیاستی زیر را برای مقامات ایرانی در جهت تدوین سیاست‌های صحیح برای بخش بانکی ارائه کرد: سیاست‌گذاران ایرانی باید اطمینان حاصل کنند که بخش بانکی تقویت شده و محیط رقابتی بازار بهبود یافته و تمرکز بازار کاهش یابد. مشخص شده است که افزایش رقابت از طریق اقداماتی مانند کاهش موانع ورود، تشویق به ارائه محصولات و خدمات جدید توسط بانک‌ها و اجرای قوانین رقابت می‌تواند انعطاف‌پذیری مالی را افزایش دهد. همچنین، واضح است که بهبود شرایط اقتصادی و ثبات برای پایداری استقلال مالی بانک‌ها ضروری است.براساس تحقیق حاضر، می‌توان بیان کرد که افزایش ثبات بانک‌ها و ارتقای رقابت در سیستم بانکی، راه‌های اصلی برای افزایش انعطاف‌پذیری مالی بانک‌های ایرانی هستند. این امر نه‌تنها تاب‌آوری بانک‌ها در برابر نوسانات اقتصادی را افزایش می‌دهد، بلکه به ثبات اقتصادی کلی و رشد اقتصاد ایران نیز کمک می‌کند.طبقه‌بندی JEL: G21، G28 ، L11 ، O15 ، L13.

Azmine Toushik Wasi, Raima Islam, Mst Rafia Islam et al.

While Large Language Models (LLM) have created a massive technological impact in the past decade, allowing for human-enabled applications, they can produce output that contains stereotypes and biases, especially when using low-resource languages. This can be of great ethical concern when dealing with sensitive topics such as religion. As a means toward making LLMS more fair, we explore bias from a religious perspective in Bengali, focusing specifically on two main religious dialects: Hindu and Muslim-majority dialects. Here, we perform different experiments and audit showing the comparative analysis of different sentences using three commonly used LLMs: ChatGPT, Gemini, and Microsoft Copilot, pertaining to the Hindu and Muslim dialects of specific words and showcasing which ones catch the social biases and which do not. Furthermore, we analyze our findings and relate them to potential reasons and evaluation perspectives, considering their global impact with over 300 million speakers worldwide. With this work, we hope to establish the rigor for creating more fairness in LLMs, as these are widely used as creative writing agents.

Melanie Subbiah, Faisal Ladhak, Akankshya Mishra et al.

Human evaluation has been the gold standard for checking faithfulness in abstractive summarization. However, with a challenging source domain like narrative, multiple annotators can agree a summary is faithful, while missing details that are obvious errors only once pointed out. We therefore introduce a new dataset, STORYSUMM, comprising LLM summaries of short stories with localized faithfulness labels and error explanations. This benchmark is for evaluation methods, testing whether a given method can detect challenging inconsistencies. Using this dataset, we first show that any one human annotation protocol is likely to miss inconsistencies, and we advocate for pursuing a range of methods when establishing ground truth for a summarization dataset. We finally test recent automatic metrics and find that none of them achieve more than 70% balanced accuracy on this task, demonstrating that it is a challenging benchmark for future work in faithfulness evaluation.

David Wan, Jesse Vig, Mohit Bansal et al.

Large Language Models (LLMs) often exhibit positional bias in long-context settings, under-attending to information in the middle of inputs. We investigate the presence of this bias in long-form summarization, its impact on faithfulness, and various techniques to mitigate this bias. To consistently evaluate faithfulness, we first compile a benchmark of eight human-annotated long-form summarization datasets and perform a meta-evaluation of faithfulness metrics. We show that LLM-based faithfulness metrics, though effective with full-context inputs, remain sensitive to document order, indicating positional bias. Analyzing LLM-generated summaries across six datasets, we find a "U-shaped" trend in faithfulness, where LLMs faithfully summarize the beginning and end of documents but neglect middle content. Perturbing document order similarly reveals models are less faithful when important documents are placed in the middle of the input. We find that this behavior is partly due to shifting focus with context length: as context increases, summaries become less faithful, but beyond a certain length, faithfulness improves as the model focuses on the end. Finally, we experiment with different generation techniques to reduce positional bias and find that prompting techniques effectively direct model attention to specific positions, whereas more sophisticated approaches offer limited improvements. Our data and code are available in https://github.com/meetdavidwan/longformfact.

Iman Islam, Esther Puyol-Antón, Bram Ruijsink et al.

Echocardiography (echo) is the first imaging modality used when assessing cardiac function. The measurement of functional biomarkers from echo relies upon the segmentation of cardiac structures and deep learning models have been proposed to automate the segmentation process. However, in order to translate these tools to widespread clinical use it is important that the segmentation models are robust to a wide variety of images (e.g. acquired from different scanners, by operators with different levels of expertise etc.). To achieve this level of robustness it is necessary that the models are trained with multiple diverse datasets. A significant challenge faced when training with multiple diverse datasets is the variation in label presence, i.e. the combined data are often partially-labelled. Adaptations of the cross entropy loss function have been proposed to deal with partially labelled data. In this paper we show that training naively with such a loss function and multiple diverse datasets can lead to a form of shortcut learning, where the model associates label presence with domain characteristics, leading to a drop in performance. To address this problem, we propose a novel label dropout scheme to break the link between domain characteristics and the presence or absence of labels. We demonstrate that label dropout improves echo segmentation Dice score by 62% and 25% on two cardiac structures when training using multiple diverse partially labelled datasets.

Stuart Margolis, Benjamin Steinberg

In this paper we compute the minimum degree of a faithful representation by partial transformations of a finite semigroup admitting a faithful completely reducible matrix representation over the field of complex numbers. This includes all inverse semigroups, and hence our results generalize earlier results of Easdown and Schein on the minimal faithful degree of an inverse semigroup. It also includes well-studied monoids like full matrix monoids over finite fields and the monoid of binary relations (i.e., matrices over the Boolean semiring). Our answer reduces the computation to considerations of permutation representations of maximal subgroups that are faithful when restricted to distinguished normal subgroups. This is analogous to (and inspired by) recent results of the second author on the minimal number of irreducible constituents in a faithful completely reducible complex matrix representation of a finite semigroup that admits one. To illustrate what happens when a finite semigroup does not admit a faithful completely reducible representation, we compute the minimum faithful degree of the opposite monoid of the full transformation monoid.

Noza Aflisia, Afrial Afrial, Asri Karolina

Not infrequently humans ignore the obligations that have been imposed on them. Assuming obligations are just routine. In this paper, the author describes the concept of human obligations, and the implications of human obligations that Allah has determined in the Quran and Hadith for Islamic education. This research is library research by reviewing and analyzing the verses of the Quran, hadith, books, proceeding, and journals related to the concept of human obligation and its implications for Islamic education. The data collection technique is to review the data sources. The data obtained is then processed with data analysis techniques through the stages of analyzing the verses of the Quran and Hadith related to human obligations so as to find the concept of human obligations, understand and analyze the characteristics of Islamic education, analyze and find implications of the concept of human obligations towards Islamic education, and draw conclusions. Through this paper, it can be seen that humans have obligations to God, fellow humans, animals, and nature. To carry out all these human obligations, humans need direction and guidance from Islamic education which always encourages humans to fulfill their obligations properly according to what Islam wants, both in knowing Allah the Creator, forming noble character, and professionalism in managing nature.

M. T. Islam, M. A. S. Akanda, F. Yesmin et al.

We investigate the magnetic domain wall (DW) dynamics in uniaxial/biaxial nanowires under a thermal gradient (TG). The findings reveal that the DW propagates toward the hotter region in both nanowires. The main physics of such observations is the magnonic angular momentum transfer to the DW. The hard (shape) anisotropy exists in biaxial nanowire, which contributes an additional torque, hence DW speed is larger than that in uniaxial nanowire. With lower damping, the DW velocity is smaller and DW velocity increases with damping which is opposite to usual expectation. To explain this, it is predicted that there is a probability to form the standing spin-waves (which do not carry net energy/momentum) together with travelling spin-waves if the propagation length of thermally-generated spin-waves is larger than the nanowire length. For larger-damping, DW decreases with damping since the magnon propagation length decreases. Therefore, the above findings might be useful in realizing the spintronic (racetrack memory) devices.

Qing Lyu, Marianna Apidianaki, Chris Callison-Burch

End-to-end neural Natural Language Processing (NLP) models are notoriously difficult to understand. This has given rise to numerous efforts towards model explainability in recent years. One desideratum of model explanation is faithfulness, i.e. an explanation should accurately represent the reasoning process behind the model's prediction. In this survey, we review over 110 model explanation methods in NLP through the lens of faithfulness. We first discuss the definition and evaluation of faithfulness, as well as its significance for explainability. We then introduce recent advances in faithful explanation, grouping existing approaches into five categories: similarity-based methods, analysis of model-internal structures, backpropagation-based methods, counterfactual intervention, and self-explanatory models. For each category, we synthesize its representative studies, strengths, and weaknesses. Finally, we summarize their common virtues and remaining challenges, and reflect on future work directions towards faithful explainability in NLP.

Han Meng, Xiaosong He, Zexing Chen et al.

Some Natural Language Generation (NLG) tasks require both faithfulness and diversity. The decoding strategy is intensively related to the quality of the generated text. Strategies such as beam search, greedy search, etc., perform with low diversity and high repetition. On the other hand, guided decoding, the solution towards diversity, may generate unfaithful expressions. To this end, this paper presents Information Filter upon Diversity-Improved Decoding (IFDID) to obtain the tradeoff between diversity and faithfulness. IFDID is a two-stage decoding strategy leveraging the proposed Enhance-Filter framework, which achieves the tradeoff by increasing the probabilities of some typical tokens being selected and subsequently filtering them by their information amount. To verify the effectiveness, we compare our method with other baselines on related CommonGEN, RocStories and AdGen benchmarks, which cover Chinese and English datasets. Our numerical experimental results and human evaluation outcomes verify the effectiveness of the proposed approach, as our approach achieves a 1.24 higher ROUGE score describing faithfulness as well as higher diversity represented by 62.5% higher upon Dist-2 than traditional approaches, demonstrating that IFDID is a novel SOTA decoding strategy for the tradeoff between diversity and faithfulness.

Zainul Mun'im

Ihya’ ‘Ulum al-Din is one of the most studied books in the pesantren, especially in the pesantren of Sidogiri. The main reason is because the book has been considered as a book of sufism with high values of moderation. This study aims to identify the moderation values of da'wah in the Ihya’ and examine its effect on the da'wah paradigm of pesantren of Sidogiri. This study uses Gadamer's hermeneutic approach to connect the theory in the ihya' and the reality of da'wah at the pesantren of Sidogiri. Data obtained through documentation, interviews and observations. This study shows that the concept of da'wah in the Ihya’  is full of high values of religious moderation. This can be understood from the concept of amr ma'ruf nahi munkar which al-Ghazali explained in his work. This research also shows that the moderation value in the Ihya’ affects the da’wah paradigm of pesantren of Sidogiri which is moderate and full of persuasive approaches.

Rini Hayati Lubis

This study aims to analyze the determining factors in paying zakat to BAZNAS (Badan Amil Zakat Nasional) Mandailing Natal Regency, where the average community in the area is Muslim which creates a respectable opportunity for BAZNAS institutions to collect zakat from muzakki (person who obliged to pay zakat) yet in fact zakat collection decreased in 2019. The variables that may cause the decline of zakat collection include trust, religiosity and income. The method used in this research are several tests of data analysis techniques such as validity test, reliability test, classic assumption test, hypothesis test, and multiple linear regression analysis using SPSS application. This research is a quantitative research occupying quota sampling technique so that the sample size is 100 muzakki. The results of the study showed that trust has no effect on the interest in paying zakat, because zakat management in BAZNAS Mandailing Natal Regency considered non optimal in supervision. Meanwhile, the variables of religiosity and income have a significant effect. This means that the higher level of religious belief and income will affect higher interest in paying zakat. Tujuan penelitian ini untuk menganalisis faktor-faktor penentu dalam membayar zakat di BAZNAS Kab Mandailing Natal, dimana rata-rata masyarakat mandaling sekitar 80 persen adalah mayoritas muslim dan ini menjadikan suatu peluang yang cukup besar bagi  lembaga BAZNAS, akan tetapi pada kenyatan dana zakat mengalami penurunan di tahun 2019, variabel yang menyebabkan penurunan antara lain kepercayaan, religiusitas, dan pendapatan. Metode dalam penelitian adalah menggunakan beberapa teknik analisis data seperti uji, validitas, reabilitas, asumsi klasik, uji hipotesis, dan uji analisis regresi linear berganda dengan menggunakan aplikasi SPSS. Jenis penelitian ini yaitu kuanlitatif, dengan teknik pengambil sampel yaitu quota sampling sehingga jumlah sampel yaitu 100 muzzaki (wajib membayar zakat). Hasil dalam penelitian menunjukan bahwa kepercayaan tidak berpengaruh terhadap minat membayar zakat, disebabkan masih belum optimalnya dalam pengelolaan manajemen zakat di BAZNAS Kabupaten Mandailing Natal. Sedang untuk variabel religius dan pendapat berpengaruh secara signifikan. Artinya semakin tingkat religius dan pendapatn akan mempengaruhi minat dalam membayar zakat.

Alek Alek, Abdul Gafur Marzuki, Muhammad Farkhan et al.

Self-assessment is one of alternative ways to evaluate students’ speaking talent in English. Through this evaluation, students are allowed to discover, know, and develop their speaking skill. Nonetheless, this sort of project was probably not common in Indonesia. Thus, this study was aimed to know students’ perception of using self-assessment for assessing their oral performance at Link and Match vocational high school. The information of this study was gathered by means of questionnaire which consists of 5 questions about the use of self-assessment. In this qualitative study, the data had been analyzed descriptively. There have been 30 vocational high school students who stand in Multimedia Major as the participant of this study. The results of this study indicated that most of the students thought that self-assessment very helpful for them because it allowed them to know their functionality and develop it to achieve the course goal specifically the speaking course objective. However, some students though that self-assessment was considerably useful since the teacher not often use this assignment and the students not take pleasure in whereas attempting to assess themselves. Self-assessment is very useful in exploring and assessing students speaking skill.