Yönetim Bilişim Sistemleri Bölümü Koleksiyonu
Bu koleksiyon için kalıcı URI
Güncel Gönderiler
Öğe Continuous intuitionistic Fuzzy AHP & CODAS methodology for automation degree selection(Old City Publishing Inc, 2024) Alkan, Nursah; Otay, Irem; Gul, Alize Yaprak; Demir, Zeynep Burcu Kizilkan; Doğan, OnurThe automotive industry's evolution thrives on technological innovation, prioritizing efficiency, safety, and sustainability. Recent improvements in autonomous driving and IoT integration have revolutionized vehicle design, safety, and maintenance with different automation degrees from partial human control to full automation. Selecting these automation degrees involves complicated Multi-Criteria Decision-Making (MCDM) encompassing technical feasibility, societal impact, and regulatory compliance. Utilizing Analytic Hierarchy Process (AHP) and Combinative Distance-Based Assessment (CODAS) offers a structured framework to navigate these complexities. AHP establishes criteria importance, while CODAS handles uncertainties, enabling informed decisions balancing technology with ethical, societal, and regulatory considerations. Fuzzy extensions further refine these methodologies, empowering the industry to adeptly address subjective perceptions and ambiguous data, enhancing the decision-making framework for automotive technology evolution. This paper navigates the intricate landscape of automation degree selection within the automotive industry evolution, employing a structured approach merging fuzzy AHP and fuzzy CODAS methods by utilizing Continuous Intuitionistic Fuzzy Set (CINFUS). This approach not only brings a new perspective to autonomous vehicles but also highlights the importance of choosing the right automation degree. Moreover, a sensitivity analysis involved adjusting the weights assigned to different criteria within the Continuous Intuitionistic Fuzzy (CINFU) AHP framework. By systematically altering these weights and observing their impact on the final automation degree selection, decision-makers can understand the sensitivity of the chosen automation degree to changes in priority among criteria.Öğe Illustration of customer analytics in public procurement(Ankara Yıldırım Beyazıt Üniversitesi, 2021) Uysal, Ümit Cengiz; Hızıroğlu, Abdulkadir; Karabacak, Muhammed EminAccording to European Commission and OECD, the share of public procurement in national economies (GDP) is 14% in EU countries and 12% in OECD countries. This rate is an important policy tool in Turkey is 7%. Public procurement is essentially made under the Public Procurement Act 4734. Besides, exceptions have been arranged for some institutions and organizations in order to meet the needs quickly and on-site. At this point, the State Supply Office appears as a central purchasing institution based on the role of intermediary between suppliers and public institutions. The aim of the study is to determine the profiles and purchasing tendencies of the customers who buy from the State Material Office by catalog method within the framework of customer analytics. In this context, customer segments based on value and behavior were created through analytical marketing methods RFM (recency, frequency and monetary) analysis. A strategy map was determined with the results obtained and the results were monitored on the business intelligence platform. Customer analytics are used extensively by the leading companies of the banking, telecom and retail sectors and significant outputs are achieved. Within this framework, customer analytical studies conducted in the public market are also important.Öğe A comparative analysis of machine learning models for time prediction in food delivery operations(İzmir Bakırçay Üniversitesi, 2024) Yalçınkaya, Elmas; Hızıroğlu, Ourania AretaAccurate time estimation is crucial for ensuring customer satisfaction and operational efficiency in the growing food delivery sector. This paper focuses on comprehensively analyzing factors affecting food delivery times and assessing the effectiveness of machine learning models in forecasting delivery times. For this purpose, authors incorporated a detailed dataset from a food delivery company of the Kaggle platform, encompassing delivery address, order time, delivery time, weather conditions, traffic intensity, and delivery person's profile information. The study evaluated the effectiveness and performance of various machine learning models such as Linear Regression, Decision Trees, Random Forests, and particularly XGBRegressor, using metrics like MAE, RMSE, and R². The results demonstrate that ensemble methods— XGBRegressor—outperformed models in accurately predicting delivery times. Additionally, a thorough analysis of feature importance uncovered the factors influencing delivery time estimation. This study offers insights into leveraging machine learning techniques to optimize food delivery operations and enhance customer satisfaction. The discoveries can assist food delivery platforms in deploying effective time estimation models and emphasizing factors for predictionsÖğe Classification of breast cancer using artificial neural network algorithms(2021) Ölmez, Emre; Hızıroğlu, Ourania Areta; Er, OrhanBreast cancer is a malignant tumor that has developed from cells of the breast. Breast cancer is one of the most fatal diseases in the world and a relatively common cancer in Turkey. Breast cancer diagnosis has been approached by various machine learning techniques for many years. In this study, two different probabilistic neural network (PNN) structures were used for breast cancer’s diagnosis. The PNN results were compared with the results of the multilayer, learning vector quantization neural networks and the results of the previous reported studies focusing on breast cancer’s diagnosis and using the same dataset. It was observed that the PNN is the best classification accuracy with 98.10% accuracy obtained via 3-fold cross validation. The present paper describes how this technique can be applied to the breast tissue classification and the breast cancer detection for medical devices. The purpose of this study is the classification of the variability of impedivity observed in normal and pathological breast tissue.Öğe An application on Chest X-Ray images for the detection of tuberculosis disease by employing deep convolutional neural networks(İzmir Bakırçay Üniversitesi, 2023) Koç, Hatice; Hızıroğlu, Abdulkadir; Erbaycu, Ahmet EminTuberculosis is the second infectious disease causing death after COVID-19. Diagnosing it is an easy and cheap via chest radiographs. However, some countries lack medical personnel and equipment for tuberculosis detection on chest radiographs. Computer-aided diagnosis and computer-aided detection systems utilizing deep learning can be employed to identify tuberculosis on medical images. Although there are some studies, they are insufficient for unbiased systems because these systems require the datasets having different features. The aim of this study is to evaluate the performance of pretrained networks for a classification application on chest X-ray images by utilizing the dataset from the Hospital in Turkey and Montgomery Count Dataset. The predictive models were implemented with the pre-trained DCNNs such as ResNet-50, Xception, and GoogLeNet. An Xception model provides the best performance.Öğe Artificial intelligence in healthcare industry: A transformation from model-driven to knowledge-driven DSS(2022) Hızıroğlu, Abdulkadir; Pişirgen, Ali; Özcan, Mert; İlter, Halil KemalHealthcare professionals and inter (or multi) disciplinary academia have been paying more attention to decision support systems (DSS) for improved decision making during their health service processes or management, as well as clinical practices. Although there have been numerous DSS applications in the healthcare industry, it has been intended to provide a categorical snapshot view of current implementations or academic work at specific DSS types for better understanding the application domains by addressing the gap in the literature. To achieve this, it has been focused on DSS applications in healthcare specifically by concentrating on two main types: model-driven and knowledge-driven. In this context, relevant information systems and medical science literatures were reviewed. For health service problems like hospital placement decisions and homecare route planning, model-driven DSS applications are used for optimization and modelling. Both conventional operations research techniques like optimization, decision analysis, simulation, and multi-criteria decision making, as well as contemporary ones like heuristic search, benefit from these applications. In addition, artificial intelligence techniques help health decision makers via knowledge-driven DSS applications, specifically clinical decision support systems (CDSS). Artificial intelligence applications can also assist health professionals in enhancing their decision-making abilities by incorporating complex operational rules and developing such procedures as single-or multi-agent systems. This research focuses on what to emphasis on while designing a DSS in the healthcare setting, such as which programming or modelling languages to employ and how to transform a model-driven DSS into a knowledgedriven DSS, or how to create the DSS more intelligent. Overall, this study indicates a present course for DSS and offers useful knowledge for both scholars and professionals in the healthcare domain.Öğe Sentiment analysis in Turkish tweets using different machine learning algorithms(İzmir Bakırçay Üniversitesi, 2024) Avvad, Hunaıda; Ereren, EcemUnderstanding emotions in any written text is considered as a hot topic for many researchers in the field of text mining, especially with the large contribution of users over the web 2.0 and with the growth of the different social media platforms. In this study we analysed emotions on Turkish text and studied the sentiment within each document using Sentiment Analysis techniques. Sentiment Analysis is the process of identifying and evaluating the emotional states contained in texts. This study aimed to investigate the effect and accuracy rate of sentiment analysis in Turkish texts. Sentiment analysis is an important field of research that helps to obtain important data in many areas such as marketing, social media analysis, and customer feedback. A comprehensive data set consisting of Turkish tweets from Kaggle was used and the emotional states of the texts were labelled. This data set consists of a variety of tweets with different topics and emotional tones. Using natural language processing techniques and machine learning algorithms, the data set was processed, and the model was trained. Within the scope of the study, different root extraction methods and a vector space model were used. In addition, machine learning algorithms such as Naive Bayes, Random Forest, Decision Tree, Gradient Boosting, Bernoulli Naive Bayes, Logistic Regression, K-Neighbours-Classifier, and Support Vector Classifier were applied to evaluate accuracy. This study aims to emphasize the importance of sentiment analysis in Turkish texts, to examine the impact of the methods used and to form a basis for future studies.Öğe Comparing the Performance of Ensemble Methods in Predicting Emergency Department Admissions Using Machine Learning Techniques(İzmir Bakırçay Üniversitesi, 2024) Yapıcı, Murat Emre; Hızıroğlu, A. Kadir; Erdoğan, Ali MertHealthcare data collection, storage, retrieval, and analysis are enabled by various technologies and tools in health information systems. These systems include health information exchanges, telemedicine platforms, clinical decision support systems, and electronic health records. They aim to improve patient outcomes, provider communication, and healthcare workflows. Machine learning is being used in emergency rooms to address challenges such as increasing patient volume, limited resources, and the need for quick decisions. Machine learning algorithms can assist in triage and risk stratification by identifying patients requiring urgent care and predicting the severity of their condition. By analyzing various patient data sources, machine learning can detect patterns and indicators that human clinicians may miss, enabling early intervention and potentially saving lives. However, there is a lack of comparative evaluation of ensemble methods used in analysis. Therefore, this study aims to thoroughly examine and analyze various ensemble methods to understand their efficacy and performance, contributing valuable insights to researchers and practitioners.Öğe Prediction of lung cancer with fuzzy logic methods: A systematic review(İzmir Bakırçay Üniversitesi, 2024) Aslan, Beyza; Hızıroğlu, Ouranıa AretaAccording to the World Health Organization (WHO), lung cancer is the primary cause of cancer-related deaths worldwide and is known to have the highest mortality rate among both men and women. Early and accurate detection of lung cancer can lead to better treatments and outcomes. Different methods can be used to diagnose a complex and uncertain disease, such as lung cancer, and fuzzy logic is one of these methods. The challenge of diagnosing lung cancer nodules, coupled with the high mortality rate of lung cancer, underscores the significance of using fuzzy logic. Fuzzy logic offers a problem-solving approach that relies on logical rules and if-then statements, incorporating human experience. There are many studies in the literature on the diagnosis of lung cancer with fuzzy logic approaches, and it is important to examine these studies to provide a general framework on this subject. Therefore, this systematic review aims to synthesize and evaluate the current evidence on the application of fuzzy logic methods in lung cancer prediction and diagnosis, and thus can provide a guide to researchers and decision makers who want to work in this field. The study followed the PRISMA guidelines for systematic reviews, ensuring a structured and transparent approach to the research process. Scopus, Web of Science (WoS), PubMed, and IEEE Explore databases were searched to find relevant studies, and appropriate studies were carefully reviewed. The inclusion and exclusion criteria were clearly defined, and the analysis process was performed independently. Out of 222 initially identified studies, 51 met the inclusion criteria and were analyzed in depth. The most commonly used fuzzy logic techniques were Fuzzy Rule-Based Systems, Fuzzy C-Means Clustering, and Fuzzy Inference Systems. Studies reported accuracy rates ranging from 85% to 98% in lung cancer prediction and diagnosis. Hybrid models combining fuzzy logic with other machine learning techniques showed particularly promising results. Fuzzy logic methods demonstrate significant potential in improving the accuracy of lung cancer prediction and diagnosis. However, further research is needed to standardize approaches and validate these methods in large-scale clinical settings. The integration of fuzzy logic with other artificial intelligence techniques presents a promising direction for future developments in lung cancer diagnostics.Öğe Şirketlerin sürdürülebilirlik web sayfalarının erişilebilirlik, kullanılabilirlik ve güvenlik perspektiflerinden kalite değerlendirmesi: Türkiye örneği(Selçuk Burak HAŞIOĞLU, 2024) Yüksel, Sıla Azer; Peker, SerhatKurumsal sürdürülebilirlik web sayfaları, uzun vadeli iş sürekliliği için çok önemli olan çevresel etkiyi azaltma stratejilerini yansıtmaktadır. Bu bağlamda, bu sayfalarda erişilebilirlik, kullanılabilirlik ve güvenlik gibi kriterlerin sağlanması oldukça önemlidir. Bu makale, kurumsal sürdürülebilirlik web sayfalarının kalitesini erişilebilirlik, kullanılabilirlik ve güvenlik perspektiflerinden değerlendirmektedir. Örneklemimiz 71 Türk şirketinden oluşmaktadır ve analiz, TAW, GTmetrix, SUCURI, Google Mobile-Friendly ve Dead Link Checker otomatik çevrimiçi test araçları kullanılarak sürdürülebilirlik web sayfaları üzerinde gerçekleştirilmiştir. Sonuçlar, Türk şirketleri arasında sürdürülebilirlik web sayfalarının genel kalitesinin artırılması gerektiğini göstermektedir. Bu araştırmanın bulguları aynı zamanda temel sorunları ele almakta ve web yöneticileri ve geliştiricileri için yapıcı bir geri bildirim işlevi görerek onları bu web sayfalarının erişilebilirlik, kullanılabilirlik ve güvenlik yönlerinde iyileştirmeler yapmaya yönlendirmektedir. Dolayısıyla bu araştırma, söz konusu web sayfalarının performansına ilişkin değerli bilgiler sunmayı ve böylece kurumsal sürdürülebilirlik iletişimi ve şeffaflığının ilerlemesine katkıda bulunmayı amaçlamaktadır.Öğe The effect of the COVID-19 pandemic on the perceived stress levels and psychological resilience of healthcare professionals(MediHealth Academy Yayıncılık, 2023) Kaynak, Kezban Özçelik; Öztuna, BarışAim: It is aimed to contribute to the literature with the broad support of participants actively working in the field during the COVID-19 pandemic in Turkey. This study was conducted to examine the effect of the COVID-19 pandemic on the perceived stress levels and psychological resilience of healthcare professionals.Material and Method: A total of 856 healthcare professionals, actively working in the COVID-19 pandemic process across Turkey, participated in the research. The data in the study were collected using the “11 Demographic Questions”, the “Four-Item Perceived Stress Scale”, developed by Cohen and friends, and the “Six-Item Brief Resilience Scale”, developed by Smith and friends to measure psychological resilience levels. The statistical analysis of the study was performed by using SPSS 23. The data, which were not normally distributed, were compared using the Mann Whitney U test and the Kruskal Wallis test. Correlation between the variables was examined via Spearman’s correlation analysis and the data, which were not normally distributed, were presented as median.Results: It was found that the mean score of the perceived stress scale was 12.7±2.9 and the mean score of psychological resilience was 17.8±4.9. It was determined that there was a moderately negative significant correlation between perceived stress and psychological resilience (r:-0.542 p:<0.001).Conclusion: The results suggested that COVID-19, whose impacts have been felt globally, increased the stress level of healthcare professionals and decreased their psychological resilience.Öğe Türkiye'ye yönelik dış turizm talebi açısından ülkelerin kümeleme analizi ile sınıflandırılması(Osman SAĞDIÇ, 2022) Aydoğdu Ulukan, Ece; Peker, SerhatTurizm ülkelerin ekonomik gelişimi için en önemli unsurlardan biridir. Ülkelere gelen yabancı turistlerin verilerinin analiz edilmesi bu gelişime katkı sağlaması açısından büyük önem taşısa da uluslararası turizme yönelik Türkiye’de yeteri kadar çalışma bulunmamaktadır. Bu çalışmanın amacı, farklı ülkelerden Türkiye’ye olan dış turizm talebini kümeleme analizi kullanarak incelemek ve Türkiye’ye turist gönderen bu ülkeleri sınıflandırmaktır. Bu bağlamda, ülkelerin gelir düzeyleri, ülkelerden çıkan turist sayıları, çıkan turist sayılarında Türkiye’nin payı ve turistlerin Türkiye’de konaklama süresi gibi faktörler dikkate alınmış ve iki aşamalı kümeleme yöntemi kullanılarak ülkeler gruplandırılmıştır. Elde edilen ülke grupları, kullanılan değişkenler ışığında karakterize edilmiştir. Bu çalışma sonucunda oluşturulan ülke profillerinin, politika yapıcılarının etkin stratejiler geliştirmesinde yardımcı olacağına inanılmaktadır.Öğe In text classification, bitcoin prices and analysis of expectations in social media with artificial neural networks(Burdur Mehmet Akif Ersoy University, 2020) Çılgın, Cihan; Ünal, Ceyda; Alıcı, Serkan; Akkol, Ekin; Gökşen, YılmazIn recent years, Web 2.0 services such as blogs, tweets, forums, emails have been widely used as communication channel. Also, social media; it is considered to be the easiest and most up-to-date way to both share information and express opinions such as requests, complaints, and wishes. As in many fields, the effect of social media on Bitcoin prices has been addressed in the last few years. Bitcoin is an investment tool that has been underlined for years, and is increasing in popularity day by day. Bitcoin, an electronic currency system that is decentralized, states a radical change in financial systems that has attracted many users. In this study, interaction of social media with Bitcoin price was revealed, particularly based on tweets obtained from Twitter channel. For this purpose, various analyses were carried out by using classification algorithms in machine learning methods over a total of 2,819,784 tweets posted by Twitter users between 06.10.2018-19.05.2019. When the findings were evaluated, Artificial Neural Networks with the highest accuracy rate of 90% was used in text classification. In addition, bilateral correlations were made with Bitcoin prices and classified positive / negative tweet rates. The correlation coefficient of 0.681 was found to be positively correlated with higher than moderate strength.Öğe Müşteri şikâyet yönetiminde firmaların performanslarının değerlendirilmesi: Kümeleme analizi incelemesi(Dicle Üniversitesi, 2022) Ödev, Gamze; Peker, SerhatMüşteri memnuniyetinde, hizmet ve ürünün kalitesi kadar müşteri şikayetlerinin dikkate alınması ve etkili bir şekilde yönetilmesi de oldukça önemli rol oynar. Günümüzde online ortamlarda şikayet daha fazla tercih edilmektedir. Bu çalışmanın amacı, kümeleme analizini kullanarak internet ortamında firmaların aldığı müşteri şikayetlerini ve bunları yönetim performanslarını değerlendirmektir. Bu amaca yönelik Sikayetvar.com internet sitesinden elde edilen veriler, CRISP-DM (Cross Industry Standard Process for Data Mining; Çapraz Endüstri Veri Madenciliği Standart Süreci) adımları baz alınarak iki aşamalı kümele analizi yöntemiyle analiz edilmiş ve elde edilen firma kümeleri profillenmiştir. Ayrıca elde edilen sonuçlar sektör bazlı olarak değerlendirilmiştir. Bu çalışmada önerilen yaklaşım ile firmalar şikayet yönetim performanslarını tespit edebilecek, diğer firmalar içindeki yerini görebilecek ve bu bağlamda başarılı firma profillerini baz alarak kendilerini geliştirebileceklerdir.Öğe Creating a comprehensive data set for deception detection studies in Turkish texts(Suat TEKER, 2024) Akkol, Ekin; Gökşen, YılmazPurpose- Deception detection has gained increasing importance with the widespread use of digital communication and online platforms. While numerous studies have been conducted on deception detection in various languages, a significant gap remains in the availability of a Turkish-language dataset for detecting deceptive reviews. This study addresses this gap by creating a comprehensive dataset specifically for deception detection in Turkish hotel reviews, including real, fake, and AI-generated comments. The dataset aims to facilitate research on deception detection, enhance the reliability of user-generated content, and contribute to the development of automated methods for identifying deceptive texts. Methodology- The study included a dataset of 5,013 Turkish hotel reviews, including real reviews from Tripadvisor, fake reviews generated by humans, and fake reviews generated by AI using the OpenAI GPT API. The collected dataset underwent extensive preprocessing to ensure quality and reliability, including data cleaning, filtering criteria, and balancing the distribution of real and fake comments. Descriptive and statistical analyses were performed to identify linguistic patterns and structural differences across these three categories. Specifically, linguistic features such as comment length, complexity, readability (measured using the Gunning Fog Index), and pronoun usage were examined. Findings- Real comments are longer and more detailed than fake and AI-generated comments, while fake comments are simpler and clearer, which supports deception detection studies in other languages. AI-generated comments frequently use the pronoun ‘we’, while fake comments tend to mimic personal experience with the pronoun ‘I’. In addition, the pronoun usage in real comments is more balanced and shows an authentic language structure. Conclusion- This study makes important contributions for fake comment detection by providing the first large-scale Turkish deception detection dataset. The findings can help businesses improve the credibility of online comments. Future work could focus on machine learning applications and comparisons with different languages.Öğe Business analytics in customer lifetime value: an overview analysis(Wiley Periodicals, Inc, 2025) Doğan, Onur; Hızıroğlu, Abdulkadir; Pisirgen, Ali; Seymen, Omer FarukIn customer-oriented systems, customer lifetime value (CLV) has been of significant importance for academia and marketing practitioners, especially within the scope of analytical modeling. CLV is a critical approach to managing and organizing a company's profitability. With the vast availability of consumer data, business analytics (BA) tools and approaches, alongside CLV models, have been applied to gain deeper insights into customer behaviors and decision-making processes. Despite the recognized importance of CLV, there is a noticeable gap in comprehensive analyses and reviews of BA techniques applied to CLV. This study aims to fill this gap by conducting a thorough survey of the state-of-the-art investigations on CLV models integrated with BA approaches, thereby contributing to a research agenda in this field. The review methodology consists of three main steps: identification of relevant studies, creating a coding plan, and ensuring coding reliability. First, relevant studies were identified using predefined keywords. Next, a coding plan-one of the study's significant contributions-was developed to evaluate these studies comprehensively. Finally, the coding plan's reliability was tested by three experts before being applied to the selected studies. Additionally, specific evaluation criteria in the coding plan were implemented to introduce new insights. This study presents exciting and valuable results from various perspectives, providing a crucial reference for academic researchers and marketing practitioners interested in the intersection of BA and CLV.Öğe Predicting the duration of professional tennis matches using MLR, CART, SVR and ANN techniques(Springer International Publishing Ag, 2024) Duen, Serdar; Peker, SerhatThis research aims to predict the duration of professional tennis matches by utilizing a dataset that includes player statistics, match characteristics and court attributes. Various machine learning techniques, such as multiple linear regression (MLR), classification and regression trees (CART), support vector regression (SVR) and artificial neural networks (ANN), are applied for this purpose. The study involves a comprehensive dataset spanning professional tournaments from 1993 to 2022. Separate predictive models were developed for tournaments played over 3 and 5 sets employing the corresponding ML techniques and their performances were compared. The findings revealed that the predictive models with MLR and SVR methods excel in best-of-3 set matches, while the ones with SVR and ANN exhibit superior performance for best-of-5 set matches. This research contributes valuable insights into the factors influencing match duration and aids in developing more effective predictive models for tennis events.Öğe Parallelism-based session creation to identify high-level activities in event log abstraction(Springer International Publishing Ag, 2024) Doğan, Onur; De Leoni, MassimilianoProcess mining utilizes event data to gain insights into the execution of processes. While techniques are valuable, their effectiveness may be hindered when dealing with highly complex processes that have a vast number of variants. Additionally, because the recorded events in information systems are at a low-level, process mining techniques may not align with the higher-level concepts understood at the business level. Without abstracting event sequences to higher-level concepts, the outcomes of process mining, such as discovering a model, can become overly complex and challenging to interpret, rendering them less useful. Some research has been conducted on event abstraction, often requiring significant domain knowledge that may not be readily accessible. Alternatively, unsupervised abstraction techniques may yield less accurate results and rely on stronger assumptions. This paper introduces a technique that addresses the challenge of limited domain knowledge by utilizing a straightforward approach. The technique involves dividing traces into batch sessions, taking into account relationships between subsequent events. Each session is then abstracted as a single high-level activity execution. This abstraction process utilizes a combination of automatic clustering and visualization methods. The proposed technique was evaluated using a randomly generated process model with high variability. The results demonstrate the significant advantages of the proposed abstraction in effectively communicating accurate knowledge to stakeholders.Öğe Enhancing hospital services: Utilizing chatbot technology for patient inquiries(Springer International Publishing Ag, 2024) Doğan, Onur; Gurcan, Omer FarukThis paper investigates the implementation and impact of a chatbot system within a hospital environment to address patient inquiries. The study focuses on the development and deployment of a non-medical knowledge-based chatbot designed to respond to a spectrum of queries typically handled by hospital call centers. The chatbot system's primary objective is to provide prompt, accurate, and accessible information to patients, streamlining their interaction process and reducing waiting times. The research explores the technical architecture and functionality of the chatbot, emphasizing its intuitive interface and ability to cater to a diverse range of patient queries. Utilizing a chatbot, data collection includes quantitative analysis of patient inquiries' pre- and post-chatbot implementation, alongside qualitative insights derived from patient feedback. The developed chatbot based on embedding techniques and a pre-trained Large Language Model gives an accuracy of 89%. Preliminary findings emphasized the potential for intelligent virtual assistants to transform patient interactions, enhance information accessibility, and ultimately contribute to the overall improvement of hospital services.Öğe Evaluating the relationship between climate change, food prices, and poverty: empirical evidence from underdeveloped countries(Springer, 2024) Acci, Yunus; Uçar, Emine; Uçar, Murat; Acci, Reyhan CafriClimate change is a critical global issue with wide-ranging impacts, particularly on agriculture. This study examines how climate change influences food prices and poverty in underdeveloped countries. Rising temperatures and extreme weather events are diminishing agricultural productivity, leading to increased food prices and worsening poverty. The research involved developing a climate change index using an autoencoder model, which can learn the important features of data and translate it into a lower-dimensional representation. This index was based on variables such as carbon emission rates, annual average rainfall, forest cover, fossil fuel consumption, renewable energy use, and temperature changes. The relationship between this climate change index and food prices and poverty was analyzed using panel causality methods. Additionally, food prices from 2020 to 2030 were projected using various time series forecasting techniques to determine the most accurate predictive model. The findings indicate that while climate change does not significantly affect poverty when considering all countries as a panel, it does have a notable impact on food prices. This underscores the need for effective policy measures to address the effects of climate change on food costs. To mitigate these impacts, it is essential for policymakers to enhance agricultural resilience through sustainable practices and targeted interventions. Future research should expand the dataset and include a broader range of countries to gain a more comprehensive understanding of how climate change affects food prices and poverty.
