Detection Method for Classifying Malicious Firmware

A malicious firmware update may prove devastating to the embedded devices both that make up the Internet of Things (IoT) and that typically lack the same security verifications now applied to full operating systems. This work converts the binary headers of 40,000 firmware examples from bytes into 1024-pixel thumbnail images to train a deep neural network. The aim is to distinguish benign and malicious variants using modern deep learning methods without needing detailed functional or forensic analysis tools. One outcome of this image conversion enables contact with the vast machine learning literature already applied to handle digit recognition (MNIST). Another result indicates that greater than 90% accurate classifications prove possible using image-based convolutional neural networks (CNN) when combined with transfer learning methods. The envisioned CNN application would intercept firmware updates before their distribution to IoT networks and score their likelihood of containing malicious variants. To explain how the model makes classification decisions, the research applies traditional statistical methods such as both single and ensembles of decision trees with identifiable pixel or byte values that contribute the malicious or benign determination.

Deep Learning for Identifying Malicious Firmware

A malicious firmware update may prove devastating to the embedded devices both that make up the Internet of Things (IoT) and alsothat typically lack the same security verifications now applied to full operating systems. This work converts the binary headers of 40,000 firmware examples from bytes into 1024-pixel thumbnail images to train a deep neural network. The aim is to distinguish benign and malicious variants using modern deep learning methods without needing detailed functional or forensic analysis tools. One outcome of this image conversion enables contact with the vast machine learning literature already applied to handle digit recognition (MNIST). Another result indicates that greater than 90% accurate classifications prove possible using image-based convolutional neural networks (CNN) when combined with transfer learning methods. The envisioned CNN application would intercept firmware updates before their distribution to IoT networks and score their likelihood of containing malicious variants.

Features of YouTube’s recommendation algorithms on the example of a channel about esports

The purpose of the study is to identify relevant effective strategies for interaction with the recommendation algorithm on the YouTube platform in the process of creating and distributing video content. The topic of the study is relevant because YouTube is an active global platform for the distribution of video content. YouTube is the most visited video viewing platform in Ukraine and one of the most visited websites in the world. Understanding the work of recommendation algorithms will allow you to adapt video content so that it reaches the widest possible target audience. The result of the study was the selection of the most effective parameters of videos on YouTube based on the analysis of statistical indicators on the PETR1K channel: relevance and novelty, attractiveness (clickability) of thumbnail images, frequency of uploads to the channel, viewers ‘reaction to previous videos, viewers’ returns, audience retention and more.

Checking the Integrity of Images with Signed Thumbnail Images

The integrity of images is an important and interesting field of research, since digital images are constantly encountered in everyday life today. The availability of image processing programs makes it possible for almost anyone to manipulate images without great effort. With the help of social media platforms, the hurdle for their distribution to a very large number of viewers has also been lowered. As a result, confidence in the integrity and authenticity of images, which was even stronger at the time of analogue photography, is dwindling. The aim of this work is to develop and investigate a concept that counteracts the lost trust and creates an opportunity to check the integrity of processed images. The concept is based on a combination of signed thumbnails and the logging of possible processing steps. We show that this combination has advantages compared to the existing approaches.

Phylotastic: Improving Access to Tree-of-Life Knowledge With Flexible, on-the-Fly Delivery of Trees

A comprehensive phylogeny of species, i.e., a tree of life, has potential uses in a variety of contexts, including research, education, and public policy. Yet, accessing the tree of life typically requires special knowledge, complex software, or long periods of training. The Phylotastic project aims make it as easy to get a phylogeny of species as it is to get driving directions from mapping software. In prior work, we presented a design for an open system to validate and manage taxon names, find phylogeny resources, extract subtrees matching a user’s taxon list, scale trees to time, and integrate related resources such as species images. Here, we report the implementation of a set of tools that together represent a robust, accessible system for on-the-fly delivery of phylogenetic knowledge. This set of tools includes a web portal to execute several customizable workflows to obtain species phylogenies (scaled by geologic time and decorated with thumbnail images); more than 30 underlying web services (accessible via a common registry); and code toolkits in R and Python (allowing others to develop custom applications using Phylotastic services). The Phylotastic system, accessible via http://www.phylotastic.org , provides a unique resource to access the current state of phylogenetic knowledge, useful for a variety of cases in which a tree extracted quickly from online resources (as distinct from a tree custom-made from character data) is sufficient, as it is for many casual uses of trees identified here.

Artificial Intelligence Imagery Analysis Fostering Big Data Analytics

In an era of accelerating digitization and advanced big data analytics, harnessing quality data and insights will enable innovative research methods and management approaches. Among others, Artificial Intelligence Imagery Analysis has recently emerged as a new method for analyzing the content of large amounts of pictorial data. In this paper, we provide background information and outline the application of Artificial Intelligence Imagery Analysis for analyzing the content of large amounts of pictorial data. We suggest that Artificial Intelligence Imagery Analysis constitutes a profound improvement over previous methods that have mostly relied on manual work by humans. In this paper, we discuss the applications of Artificial Intelligence Imagery Analysis for research and practice and provide an example of its use for research. In the case study, we employed Artificial Intelligence Imagery Analysis for decomposing and assessing thumbnail images in the context of marketing and media research and show how properly assessed and designed thumbnail images promote the consumption of online videos. We conclude the paper with a discussion on the potential of Artificial Intelligence Imagery Analysis for research and practice across disciplines.

Phylotastic: improving access to tree-of-life knowledge with flexible, on-the-fly delivery of trees

Abstract(1) A comprehensive phylogeny of species, i.e., a tree of life, has potential uses in a variety of contexts in research and education. This potential is limited if accessing the tree of life requires special knowledge, complex software, or long periods of training.(2) The Phylotastic project aims to use web-services technologies to lower the barrier for accessing phylogenetic knowledge, making it as easy to get a phylogeny of species as it is to get online driving directions. In prior work, we designed an open system of web services to validate and manage species names, find phylogeny resources, extract subtrees matching a user’s species list, calibrate them, and mash them up with images and information from online resources.(3) Here we report a publicly accessible system for on-the-fly delivery of phylogenetic knowledge, developed with user feedback on what types of functionality are considered useful by researchers and educators. The system currently consists of a web portal that implements 3 types of workflows to obtain species phylogenies (scaled by geologic time and decorated with thumbnail images); 19 underlying web services accessible via a common registry; and toolbox code in R and Python so that others can create applications that leverage these services. These resources cover most of the use-cases identified in our analysis of user needs.(4) The Phylotastic system, accessible viahttp://www.phylotastic.org, provides a unique resource to access the current state of phylogenetic knowledge, useful for a variety of cases in which a tree extracted quickly from online resources (as distinct from a tree custom-made from character data) is sufficient, as it is for many casual uses of trees identified here.