Çiğdem İnan ACI

Prof. Dr. Çiğdem İnan ACI

orcid

Research Focus & Metrics

Academic Career Length

17Years (2010-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 November 2019 in Expert Systems with Applications
Verified
Journal Article
Closed access
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
MKMurat Kaya
Murat Kaya
YMYuriy Mishchenko
Yuriy Mishchenko

2 June 2019 in Computational Intelligence and Neuroscience
Verified
Journal Article
Open access
PDF file
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
HGHakan Gülcan
Hakan Gülcan
The dragonfly algorithm (DA) is one of the optimization techniques developed in recent years. The random flying behavior of dragonflies in nature is modeled in the DA using the Levy flight mechanism (LFM). However, LFM has disadvantages such as the overflowing of the search area and interruption of random flights due to its big searching steps. In this study, an algorithm, known as the Brownian motion, is used to improve the randomization stage of the DA. The modified DA was applied to 15 single-objective and 6 multiobjective problems and then compared with the original algorithm. The modified DA provided up to 90% improvement compared to the original algorithm’s minimum point access. The modified algorithm was also applied to welded beam design, a well-known benchmark problem, and thus was able to calculate the optimum cost 20% lower.

31 July 2019 in Bilişim Teknolojileri Dergisi
Verified
Journal Article
Closed access
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
AÇAdem Çirak
Adem Çirak
Bu çalışmada, Konvolüsyonel Sinir Ağları (KSA) ve Word2Vec metodu kullanılarak Turkish Text Classification 3600 (TTC-3600) veri kümesi üzerinde metin sınıflandırma çalışması yapılmış ve aynı veri kümesi kullanılarak yapılan önceki çalışma ile kıyaslanmıştır. Çalışmada TTC-3600’ün ham ve Zemberek yazılımıyla gövdelenmiş halleri üzerinde iki farklı KSA eğitilmiş ve test edilmiştir. KSA ve Word2Vec metodu, klasik istatistiksel ve makine öğrenmesine dayalı sınıflandırma algoritmalarından daha iyi bir performans (%93,3 doğruluk) göstermiştir. Türkçe doğal dil işleme çalışmalarının azlığı ve bu alandaki özellik çıkarma yöntemlerinin limitli olması sebebiyle, kelimelerin semantik değerlerinin önceden eğitilmiş Word2Vec ağı ile sınıflandırmaya katılabilmesi KSA modellerinin doğruluk değerlerini arttırmıştır.

1 October 2019 in Turkish Journal of Engineering
Verified
Journal Article
Closed access
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
EÇEren Çürük
Eren Çürük
ESEsra Sarac
Esra Sarac
Cyberbullying has become a major problem along with the increase of communication technologies and social media become part of daily life. Cyberbullying is the use of communication tools to harass or harm a person or group. Especially for the adolescent age group, cyberbullying causes damage that is thought to be suicidal and poses a great risk. In this study, a model is developed to identify the cyberbullying actions that took place in social networks. The model investigates the effects of some text mining methods such as pre-processing, feature extraction, feature selection and classification on automatic detection of cyberbullying using datasets obtained from Formspring.me, Myspace and YouTube social network platforms. Different classifiers (i.e. multilayer perceptron (MLP), stochastic gradient descent (SGD), logistic regression and radial basis function) have been developed and the effects of feature selection algorithms (i.e. Chi2, support vector machine-recursive feature elimination (SVM-RFE), minimum redundancy maximum relevance and ReliefF) for cyberbullying detection have also been investigated. The experimental results of the study proved that SGD and MLP classifiers with 500 selected features using SVM-RFE algorithm showed the best results (F_measure value is more than 0.930) by means of classification time and accuracy.

25 April 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
NSNisa Vuran Sarı
Nisa Vuran Sarı
MAMehmet ACI
Mehmet ACI
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
As malware has become increasingly complex, advanced techniques have emerged to improve traditional detection systems. The increasing complexity of malware poses significant challenges in cybersecurity due to the inability of existing methods to understand detailed and contextual relationships in modern software behavior. Therefore, developing innovative detection frameworks that can effectively analyze and interpret these complex patterns has become critical. This work presents a novel framework integrating API call sequences and DLL information into a unified, graph-based representation to analyze malware behavior comprehensively. The proposed model generates initial embeddings using Node2Vec, which uses a random walk approach to understand structural relationships between nodes. Graph Attention Network (GAT) then enhances these initial embeddings, which utilizes attention mechanisms to incorporate contextual dependencies and enhance semantic representations. Finally, the enhanced embeddings are classified using Convolutional Neural Network (CNN) and Gated Recurrent Units (GRU)s, a custom hybrid CNN-GRU-3 deep learning-based model capable of effectively modeling sequential patterns. The dual role of GAT as a classifier and feature extractor is also analyzed to evaluate its impact on embedding quality and classification accuracy. Experimental results show that the proposed model achieves superior results with an accuracy rate of 0.9961 compared to state-of-the-art approaches such as ensemble learning and standalone GAT. This achievement highlights the framework’s ability to utilize contextual information for malware detection. The real-world dataset used provides a benchmark for future work, and this research lays a comprehensive foundation for advancing graph-based malware analysis.

4 February 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
Çiğdem İnan ACIÇiğdem İnan ACI
Çiğdem İnan ACI
Gizen Mutlu KorucuGizen Mutlu Korucu
Gizen Mutlu Korucu
MOMurat Ozen
Murat Ozen
MAMehmet ACI
Mehmet ACI
Predicting driver injury severity and identifying factors influencing crash outcomes are crucial for developing effective traffic safety measures. This study focuses on estimating driver injury severity (uninjured, injured, or killed) and determining critical factors affecting crash outcomes. A hybrid framework combining Deep Neural Networks (DNNs) and Random Forest (RF) is proposed, where a DNN extracts features and RF performs the final classification, leveraging ensemble methods. The results were compared with those of well-known methods (e.g., kNN, XGBoost), with the hybrid approach achieving the best performance (0.92 accuracy, 0.89 F1-macro, 0.91 F1-micro scores) in predicting injury severity. The results showed that crash type, vehicle type, driver fault, intersection type, season, time, and road type had the greatest impact, while factors like pavement condition and driver gender had minimal influence. To the best of our knowledge, this is the first study to combine DNN-based feature extraction with RF classification in the context of traffic injury severity prediction. The framework offers a new approach for drivers and policymakers, providing a deeper understanding of driver injury severity prediction and its underlying factors.