Yazar "Uzun-Per, Meryem" seçeneğine göre listele
Listeleniyor 1 - 10 / 10
Sayfa Başına Sonuç
Sıralama seçenekleri
Yayın Air ticket sales forecasting for B2B travel agencies(Emerald Publishing, 2026) Uzun-Per, Meryem; Akkaya, Ümit Murat; Kalkan, HabilPurpose – This study addresses the air ticket sales forecasting problem for business-to-business (B2B) travel agencies, a largely unexplored area compared to airlines and business-to-consumer agencies. To meet guarantee deposit requirements and manage financial risk exposure, B2B agencies require accurate demand estimation for budget planning and revenue management. The study proposes a hybrid deep learning forecasting model integrating recurrent neural network (RNN), long short-time memory (LSTM) and bidirectional long short-term memory (BiLSTM) architectures to capture both short-term fluctuations and long-term temporal dependencies in ticket sales data. Design/methodology/approach – Approximately eight years (from 3 August 2014 to 2 June 2022) of daily search and ticket sales data from a B2B travel consolidator agency were combined with external variables, including holidays, weather conditions, exchange rates, gross domestic product per capita and COVID-19 cases. Two datasets (with and without pandemic data) were constructed. After preprocessing and Random Forest feature selection, seasonal autoregressive integrated moving average (SARIMA), Prophet, RNN, LSTM and BiLSTM models were implemented and compared with the proposed hybrid architecture using a sliding window forecasting framework. Performance was evaluated using mean absolute percentage error (MAPE) and root mean squared error (RMSE). Findings – The proposed hybrid model achieved the highest forecasting accuracy across both datasets, outperforming statistical and individual deep learning models. It obtained MAPE values of 5.83% and 6.16% and RMSE values of 0.0379 and 0.0411, respectively. BiLSTM ranked as the second-best method, while SARIMA and Prophet showed substantially lower accuracy. Results also indicated that excluding pandemic variables did not significantly reduce predictive performance, but still, pandemic-related volatility increased the error rate of most of the methods and increased forecasting difficulty. Originality/value – This research is the first to examine ticket sales forecasting specifically for B2B travel agencies. It introduces a novel hybrid sequential deep learning architecture and evaluates the role of pandemic and macroeconomic variables in forecasting performance. The findings demonstrate the effectiveness of hybrid recurrent models for complex real-world demand forecasting problems.Yayın An NLP-driven framework for automated radiology–pathology concordance assessment in breast biopsy(MDPI Publishing, 2026) Esmerer, Emel; Nazlı, Mehmet Ali; Uzun-Per, Meryem; Gümüş Değidiben, Melike; Söyleyici, Merve; Tahir, Eren; Bal, MertBackground/Objectives: To develop and assess the feasibility of a natural language processing (NLP) framework for automated assessment of radiology-pathology concordance in breast biopsy using machine learning-based analysis of unstructured reports. Methods: This retrospective study included 766 paired radiology and pathology reports from ultrasound- or mammography-guided breast biopsies (August 2020-May 2024). Reports underwent translation, normalization, tokenization, lemmatization, and synonym expansion, followed by structured encoding of BI-RADS and pathology categories. Three models were trained: a Decision Tree, a LightGBM classifier, and a fine-tuned BioBERT model. Concordance labels were defined by multidisciplinary consensus. Performance metrics included accuracy, sensitivity, specificity, F1-score, area under the curve (AUC), and Cohen's kappa. SHapley Additive exPlanations (SHAP) analysis was used to identify influential features. Results: Among 766 cases, 707 (92.3%) were concordant and 59 (7.7%) were initially discordant. After excluding B3 lesions (n = 46), 13 true discordant cases remained (1.7%). Including B3 lesions increased clinically non-concordant or indeterminate cases from 1.7% to 7.7%, indicating that the apparent performance of the models is likely sensitive to case definition and dataset composition. BI-RADS 4a was the most common category (31.3%), and benign pathology (B2) accounted for 64.4% of biopsies. Within this dataset, LightGBM yielded the highest apparent AUC (0.999) (however, given the extremely small number of true discordant cases, this estimate is likely unstable and should be interpreted with caution), while BioBERT showed the strongest agreement with expert consensus (κ = 0.89). SHAP analysis identified clinically meaningful terms such as calcification, hypoechoic, ductal, and carcinoma as key contributors to model predictions. Given the very limited number of true discordant cases, these performance estimates are likely unstable and should be regarded as preliminary, requiring validation in larger, multi-center cohorts. Conclusions: This study presents a proof-of-concept NLP-based framework for radiology-pathology concordance assessment. The models showed promising performance in identifying potentially discordant cases; however, given the limited number of true discordant samples, these findings should be considered preliminary and require further validation in larger, multi-center datasets before clinical implementation.Yayın Big data testing framework for recommendation systems in e-science and e-commerce domains(IEEE, 2021) Uzun-Per, Meryem; Can, Ali Burak; Gürel, Ahmet Volkan; Aktaş, Mehmet S.Software testing is an important process to evaluate whether the developed software applications meet the required specifications. There is an emerging need for testing frameworks for big data software projects to ensure the quality of the big data applications and satisfy the user requirements. In this study, we propose a software testing framework that can be utilized in big data projects both in e-science and e-commerce. In particular, we design the proposed framework to test big data-based recommendation applications. To show the usability of the proposed framework, we provide a reference prototype implementation and use the prototype to test a big data recommendation application. We apply the prototype implementation to test both functional and non-functional methods of the recommendation application. The results indicate that the proposed testing framework is usable and efficient for testing the recommendation systems that use big data processing techniques.Yayın Feature selection in customer churn analysis: Case study in B2B business(Institute of Electrical and Electronics Engineers Inc., 2022) Sancar, Semanur; Uzun-Per, MeryemCustomer churn analysis is one of the machine learning applications that has become a hot topic in businesses with the developing technology. Since the performance of fore-casting algorithms is directly affected by the abundance of data, the literature for B2C businesses has been developed faster. In B2B businesses, on the other hand, since customer dynamics are slightly different and the number of customers is not as high as in B2C, data mining studies have been carried out less frequently. Within the scope of BiletBank R&D studies, it is aimed to analyze the customer loss of BiletBank, a B2B company. In line with the customer loss analysis target, the categorical features of BiletBank customers, such as the city they are in, and their periodic interactions with the BiletBank system, have been converted into a data set. In this study, the evaluation of the features in the data set was carried out to create a source for customer loss analysis. Evaluation of features has been implemented by establishing nine different models, including statistical, wrapper, and embedded methods. It is aimed that the feature importance determined as a result of this study will be used in the customer churn analysis studies to be carried out from now on.Yayın Kullanıcı ve öğe bazlı, geniş ve derin öğrenme tabanlı seyahat öneri sistemi(2023) Öz, Alihan; Uzun-Per, Meryem; Bal, MertTeknolojinin gelişmesi ile birlikte artan dijital bilgi miktarı ve internetin yaygınlaşması ile internet üzerinden ürün, hizmet, abonelik gibi ticaret işlemlerinin gerçekleştiği web sitelerinin sayısının da artması, beraberinde, müşterilere kişiselleştirilmiş ve doğru; ürün, hizmet ve abonelikleri sunmanın (önermenin) de önemini artmıştır. Müşterilere önerilerde yaygın olarak kullanılan ürün bazlı, kullanıcı tabanlı ve bu ikisinin birlikte kullanıldığı hibrit geleneksel yaklaşımlar çoğu çalışmada kullanılmaktadır. Geleneksel yaklaşımların, büyük ve seyrek veriler ile çalışma, kullanıcı ve ürün arasındaki karışık ilişkileri bulamama ve soğuk başlangıç (cold start) gibi problemlerinin üstesinden gelmek, derin ve geniş öğrenme sistemlerinin kullanımı ile mümkün olmuştur. Bu çalışma kapsamında, öncelikle derin ve geniş sinir ağlarına ve bunların seyahat öneri sistemlerindeki uygulamalarına kapsamlı bir bakış açısı sunulmuş ve en popüler öneri algoritmaları olan Google'ın Geniş ve Derin Algoritması ve Facebook'un Deep Learning Recommendation Model (DLRM) algoritmasına yer verilmiştir. Ardından, geniş ve derin öğrenme yaklaşımı ile kullanıcı ve ürün özelliklerinin kategorik olanlarının gömme işlemi uygulanarak, nümerik veriler ile modele beslendiği yeni bir seyahat öneri sistemi oluşturulmuştur. Önerilen yöntem gerçek bir seyahat acentesi şirketinin veri seti üzerinde uygulanmıştır. Sonuçta, kullanıcılara verilen en iyi beş öneride, %82.37 doğruluk oranı yakalanmıştır.Yayın A novel sequential pattern mining algorithm for large scale data sequences(Springer, 2022) Can, Ali Burak; Uzun-Per, Meryem; Aktaş, Mehmet SıddıkSequential pattern mining algorithms are unsupervised machine learning algorithms that allow finding sequential patterns on data sequences that have been put together based on a particular order. These algorithms are mostly optimized for finding sequential data sequences containing more than one element. Hence, we argue that there is a need for algorithms that are particularly optimized for data sequences that contain only one element. Within the scope of this research, we study the design and development of a novel algorithm that is optimized for data sets containing data sequences with single elements and that can detect sequential patterns with high performance. The time and memory requirements of the proposed algorithm are examined experimentally. The results show that the proposed algorithm has low running times, while it has the same accuracy results as the algorithms in the similar category in the literature. The obtained results are promising.Yayın On the big data processing algorithms for finding frequent sequences(Wiley, 2023) Can, Ali Burak; Zaval, Mounes; Uzun-Per, Meryem; Aktaş, Mehmet SıddıkSequential pattern mining algorithms extract trendy sequence appearances insideordered transactional datasets such as market basket datasets. There is a lack ofresearch employing big data processing techniques to locate frequent sequences onlarge-scale datasets. Furthermore, there is a need for optimized sequential patternmining algorithms that run on ordered one-dimensional sequences. We also observe alack of sequential pattern search studies in the literature, where the focus is centeredaround multi-dimensional data sequences. Existing approaches that deal with orderedone-dimensional datasets suffer from scalability issues as the amount of data to beanalyzed is enormous. This research investigates the big data processing techniquesused to find frequent sequences in large-scale datasets. It also proposes a scalablesequence pattern mining algorithm called Sequential Pattern Acquisition by ReducingSearch Space (SPARSS) designed for distributed data processing systems that effi-ciently handle large datasets containing sequential one-element data. It introducesa prototype implementation of SPARSS and provides information on the SPARSS’smemory and time requirements, which were calculated as part of experimental stud-ies on a real-world dataset. The results confirm our expectations and demonstrateSPARSS’s superior scalability and run-time efficiency compared to other distributedalgorithms.Yayın Scalable recommendation systems based on finding similar items and sequences(Wiley, 2022) Uzun-Per, Meryem; Gürel, Ahmet Volkan; Can, Ali Burak; Aktaş, Mehmet S.The rapid growth in the airline industry, which started in 2009, continued until the COVID-19 era, with the annual number of passengers almost doubling in 10 years. This situation has led to increased competition between airline companies, whose profitability has decreased considerably. They aimed to increase their profitability by making services like seat selection, excess baggage, Wi-Fi access optional under the name of ancillary services. To the best of our knowledge, there is no recommendation system for recommending ancillary services for airline companies. Also, to the best of our knowledge, there is no testing framework to compare recommendation algorithms considering their scalabilities and running times. In this paper, we propose a framework based on Lambda architecture for recommendation systems that run on a big data processing platform. The proposed method utilizes association rule and sequential pattern mining algorithms that are designed for big data processing platforms. To facilitate testing of the proposed method, we implement a prototype application.We conduct an experimental study on the prototype to investigate the performance of the proposed methodology using accuracy, scalability, and latency related performance metrics. The results indicate that the proposed method proves to be useful and has negligible processing overheads.Yayın SFNN: A secure and diverse recommender system through graph neural network and regularized variational autoencoder(Elsevier, 2025) Bahi, Abderaouf; Gasmi, Ibtissem; Bentrad, Sassi; Azizi, Mohamed Walid; Khantouchi, Ramzi; Uzun-Per, MeryemRecommender systems are frequently improved to filter information and provide users with the most relevant items. However, they face limitations in balancing appropriate and diverse recommendations while ensuring the security and integrity of user data. A new recommender system based on secure fusion neural network is pre sented in this paper. It guarantees data integrity and confidentiality while balancing accuracy and diversity. It integrates a graph neural network that models user-item interactions to improve accuracy, with a regularized variational autoencoder whose evidence lower bound loss function is enhanced by a diversity-promoting regu larization term that favors latent-space dispersion, thereby improving recommendation diversity. To optimize the combination of the two neural networks scores, an adaptive fusion mechanism is introduced to generate final predictions that consider diverse user preferences while maintaining relevance. Furthermore, our approach uses blockchain technology to encrypt and secure data storage, ensuring the integrity and confidentiality of users’ data. The experiments conducted on three datasets show that the proposed model can achieve an accuracy of 78.13 % with an intra-list diversity of 46.82 % for Retail Rocket dataset, an accuracy of 82.44 % with an intra-list diversity of 37.78 % for clothing dataset, and an accuracy of 86.16 % with an intra-list diversity of 47.65 % for MovieLens-1 M dataset.Yayın Testing the performance of feature selection methods for customer churn analysis: Case study in B2B business(Springer, 2023) Sancar, Semanur; Uzun-Per, Meryem; García Márquez, Fausto Pedro; Jamil, Akhtar; Eken, Süleyman; Hameed, Alaa AliChurn analysis has recently become one of the favorite topics of marketing teams with the development of machine learning models. This study aims to discover the most suitable feature selection (FS) model for churn analysis by using the databases of BiletBank, a business-to-business (B2B) company. It was found that some categorical data such as agency type and currency used by customers, along with periodic flight sales data, are also meaningful features for churn analysis in the BiletBank customer portfolio. This feature selection study in the database will be a source for future churn analysis studies.
