Organically produced spices and herbs were analyzed for determination of aflatoxin B1 (AFB1) by ELISA using immunoaffinity column. For this purpose 93 organic spices and 37 organic herbs were randomly selected from organic markets and organic shops in Turkey. AFB1 was detected in 58 organic spice and 32 organic herb samples. Among organic spice samples, the maximum value was detected in cinnamon sample (53 μg/kg). AFB1 was not detected in thyme samples. AFB1 levels of 41 organic spice samples were above the EU regulatory limit (5 μg/kg). Among organic herb samples the highest concentration of AFB1 (52.5 μg/kg) was detected in a rosehip sample. AFB1 levels of 21 organic herb samples were above the regulatory limits of the European Union. These results showed that more stringent measures must be taken for the prevention of mold contamination in the production of organic spices and herbs.

Mobile applications create their own security and privacy models through permission-based models. Some applications may request extra permissions that they do not need but may use for suspicious activities. The aim of this study is to identify those spare permissions requested and use this information in the security and privacy approach, which uses static and code analysis together and applies them to the existing datasets; then the results are compared and accuracy level is determined. Classification is made with an accuracy rate of 91.95%.

This paper presents a correction methodology for Long Short Term Memory (LSTM) based speech recognition. A strategy that validates with a reference database was developed for LSTM. It is conceptually simple but requires a large keyword database to match test templates. The correction method is based on the “most matching method” that is finding the word in which the system output is closest among the “Referenced Template Database”. Each LSTM model recognition output was corrected with the proposed new concept. Thus, system recognition performance was improved by correcting faulty outputs. The effectiveness, efficiency, and contribution of this approach to system performance were demonstrated by experiments. Tests carried out using different speech-text datasets and LSTM models yielded an average performance increase of 2.25%. With some advanced models, this ratio rises to 3.84%.

Background Technological developments have a significant effect on the development of smart devices. The use of smart devices has become widespread due to their extensive capabilities. The Android operating system is preferred in smart devices due to its open-source structure. This is the reason for its being the target of malware. The advancements in Android malware hiding and detection avoidance methods have overridden traditional malware detection methods. Methods In this study, a model employing AndroAnalyzer that uses static analysis and deep learning system is proposed. Tests were carried out with an original dataset consisting of 7,622 applications. Additional tests were conducted with machine learning techniques to compare it with the deep learning method using the obtained feature vector. Results Accuracy of 98.16% was achieved by presenting a better performance compared to traditional machine learning techniques. Values of recall, precision, and F-measure were 98.78, 99.24 and 98.90, respectively. The results showed that deep learning models using trace-based feature vectors outperform current cutting-edge technology approaches.

No abstract