Michael Fire

Chen Tsfaty, Michael Fire

Open source code is considered a common practice in modern software development. However, reusing other code allows bad actors to access a wide developers' community, hence the products that rely on it. Those attacks are categorized as supply chain attacks. Recent years saw a growing number of supply chain attacks that leverage open source during software development, relaying the download and installation procedures, whether automatic or manual. Over the years, many approaches have been invented for detecting vulnerable packages. However, it is uncommon to detect malicious code within packages. Those detection approaches can be broadly categorized as analyzes that use (dynamic) and do not use (static) code execution. Here, we introduce Malicious Source code Detection using Transformers (MSDT) algorithm. MSDT is a novel static analysis based on a deep learning method that detects real-world code injection cases to source code packages. In this study, we used MSDT and a dataset with over 600,000 different functions to embed various functions and applied a clustering algorithm to the resulting vectors, detecting the malicious functions by detecting the outliers. We evaluated MSDT's performance by conducting extensive experiments and demonstrated that our algorithm is capable of detecting functions that were injected with malicious code with precision@k values of up to 0.909.

Nadav Shalit, Michael Fire, Dima Kagan et al.

Public transport routing plays a crucial role in transit network design, ensuring a satisfactory level of service for passengers. However, current routing solutions rely on traditional operational research heuristics, which can be time-consuming to implement and lack the ability to provide quick solutions. Here, we propose a novel deep learning-based methodology for a decision support system that enables public transport (PT) planners to identify short-term route improvements rapidly. By seamlessly adjusting specific sections of routes between two stops during specific times of the day, our method effectively reduces times and enhances PT services. Leveraging diverse data sources such as GTFS and smart card data, we extract features and model the transportation network as a directed graph. Using self-supervision, we train a deep learning model for predicting lateness values for road segments. These lateness values are then utilized as edge weights in the transportation graph, enabling efficient path searching. Through evaluating the method on Tel Aviv, we are able to reduce times on more than 9\% of the routes. The improved routes included both intraurban and suburban routes showcasing a fact highlighting the model's versatility. The findings emphasize the potential of our data-driven decision support system to enhance public transport and city logistics, promoting greater efficiency and reliability in PT services.

Nofar Piterman, Tamar Makov, Michael Fire

Online social networks usage has increased significantly in the last decade and continues to grow in popularity. Multiple social platforms use volunteers as a central component. The behavior of volunteers in volunteer-based networks has been studied extensively in recent years. Here, we explore the development of volunteer-based social networks, primarily focusing on their key users' behaviors and activities. We developed two novel algorithms: the first reveals key user behavior patterns over time; the second utilizes machine learning methods to generate a forecasting model that can predict the future behavior of key users, including whether they will remain active donors or change their behavior to become mainly recipients, and vice-versa. These algorithms allowed us to analyze the factors that significantly influence behavior predictions. To evaluate our algorithms, we utilized data from over 2.4 million users on a peer-to-peer food-sharing online platform. Using our algorithm, we identified four main types of key user behavior patterns that occur over time. Moreover, we succeeded in forecasting future active donor key users and predicting the key users that would change their behavior to donors, with an accuracy of up to 89.6%. These findings provide valuable insights into the behavior of key users in volunteer-based social networks and pave the way for more effective communities-building in the future, while using the potential of machine learning for this goal.

Shmuel Horowitz, Dima Kagan, Galit Fuhrmann Alpert et al.

In recent years, video conferencing (VC) popularity has skyrocketed for a wide range of activities. As a result, the number of VC users surged sharply. The sharp increase in VC usage has been accompanied by various newly emerging privacy and security challenges. VC meetings became a target for various security attacks, such as Zoombombing. Other VC-related challenges also emerged. For example, during COVID lockdowns, educators had to teach in online environments struggling with keeping students engaged for extended periods. In parallel, the amount of available VC videos has grown exponentially. Thus, users and companies are limited in finding abnormal segments in VC meetings within the converging volumes of data. Such abnormal events that affect most meeting participants may be indicators of interesting points in time, including security attacks or other changes in meeting climate, like someone joining a meeting or sharing a dramatic content. Here, we present a novel algorithm for detecting abnormal events in VC data. We curated VC publicly available recordings, including meetings with interruptions. We analyzed the videos using our algorithm, extracting time windows where abnormal occurrences were detected. Our algorithm is a pipeline that combines multiple methods in several steps to detect users' faces in each video frame, track face locations during the meeting and generate vector representations of a facial expression for each face in each frame. Vector representations are used to monitor changes in facial expressions throughout the meeting for each participant. The overall change in meeting climate is quantified using those parameters across all participants, and translating them into event anomaly detection. This is the first open pipeline for automatically detecting anomaly events in VC meetings. Our model detects abnormal events with 92.3% precision over the collected dataset.

Daniel Azenkot, Michael Fire, Eran Ben Elia

Accurate forecasting of bus ridership (passengers numbers) is crucial for efficient management and optimization of public transport systems. Traditional forecasting models often fail to capture the unique and localized dynamics of different urban areas by treating the entire city as a single, homogeneous region. This paper introduces a novel framework that enhances bus ridership prediction by integrating a spatial clustering methodology with multi-dimensional feature analysis. The proposed framework utilizes a diverse set of data, including bus ridership data (by route number, time, and bus stop) complemented by a variety of open source data, such as spatial features (e.g., attractive destinations), meteorological conditions (e.g., temperature, rainfall), and temporal patterns (e.g., time of day, day of week). By clustering the urban area into distinct regions, based on the principle that bus stops in close proximity share similar ridership characteristics, a separate local forecasting model is trained for each of these clusters. This localized approach demonstrates an accuracy comparable to that of global models. The findings suggest that a spatially-aware, localized modeling strategy is effective for public transport prediction, paving the way for more targeted and efficient service improvements.

Michael Fire, Yitzhak Elbazis, Adi Wasenstein et al.

Large Language Models (LLMs) rapidly reshape modern life, advancing fields from healthcare to education and beyond. However, alongside their remarkable capabilities lies a significant threat: the susceptibility of these models to jailbreaking. The fundamental vulnerability of LLMs to jailbreak attacks stems from the very data they learn from. As long as this training data includes unfiltered, problematic, or 'dark' content, the models can inherently learn undesirable patterns or weaknesses that allow users to circumvent their intended safety controls. Our research identifies the growing threat posed by dark LLMs models deliberately designed without ethical guardrails or modified through jailbreak techniques. In our research, we uncovered a universal jailbreak attack that effectively compromises multiple state-of-the-art models, enabling them to answer almost any question and produce harmful outputs upon request. The main idea of our attack was published online over seven months ago. However, many of the tested LLMs were still vulnerable to this attack. Despite our responsible disclosure efforts, responses from major LLM providers were often inadequate, highlighting a concerning gap in industry practices regarding AI safety. As model training becomes more accessible and cheaper, and as open-source LLMs proliferate, the risk of widespread misuse escalates. Without decisive intervention, LLMs may continue democratizing access to dangerous knowledge, posing greater risks than anticipated.

Zahi Mizrahi, Shai Berkovitz, Nimrod Talmon et al.

Keeping track of how lawmakers vote is essential for government transparency. While many parliamentary voting records are available online, they are often difficult to interpret, making it challenging to understand legislative behavior across parliaments and predict voting outcomes. Accurate prediction of votes has several potential benefits, from simplifying parliamentary work by filtering out bills with a low chance of passing to refining proposed legislation to increase its likelihood of approval. In this study, we leverage advanced machine learning and data analysis techniques to develop a comprehensive framework for predicting parliamentary voting outcomes across multiple legislatures. We introduce the Voting Prediction Framework (VPF) - a data-driven framework designed to forecast parliamentary voting outcomes at the individual legislator level and for entire bills. VPF consists of three key components: (1) Data Collection - gathering parliamentary voting records from multiple countries using APIs, web crawlers, and structured databases; (2) Parsing and Feature Integration - processing and enriching the data with meaningful features, such as legislator seniority, and content-based characteristics of a given bill; and (3) Prediction Models - using machine learning to forecast how each parliament member will vote and whether a bill is likely to pass. The framework will be open source, enabling anyone to use or modify the framework. To evaluate VPF, we analyzed over 5 million voting records from five countries - Canada, Israel, Tunisia, the United Kingdom and the USA. Our results show that VPF achieves up to 85% precision in predicting individual votes and up to 84% accuracy in predicting overall bill outcomes. These findings highlight VPF's potential as a valuable tool for political analysis, policy research, and enhancing public access to legislative decision-making.

Gregory Koushnir, Michael Fire, Galit Fuhrmann Alpert et al.

Voice data is increasingly being used in modern digital communications, yet there is still a lack of comprehensive tools for automated voice analysis and characterization. To this end, we developed the VANPY (Voice Analysis in Python) framework for automated pre-processing, feature extraction, and classification of voice data. The VANPY is an open-source end-to-end comprehensive framework that was developed for the purpose of speaker characterization from voice data. The framework is designed with extensibility in mind, allowing for easy integration of new components and adaptation to various voice analysis applications. It currently incorporates over fifteen voice analysis components - including music/speech separation, voice activity detection, speaker embedding, vocal feature extraction, and various classification models. Four of the VANPY's components were developed in-house and integrated into the framework to extend its speaker characterization capabilities: gender classification, emotion classification, age regression, and height regression. The models demonstrate robust performance across various datasets, although not surpassing state-of-the-art performance. As a proof of concept, we demonstrate the framework's ability to extract speaker characteristics on a use-case challenge of analyzing character voices from the movie "Pulp Fiction." The results illustrate the framework's capability to extract multiple speaker characteristics, including gender, age, height, emotion type, and emotion intensity measured across three dimensions: arousal, dominance, and valence.

Koren Ishlach, Itzhak Ben-David, Michael Fire et al.

Embedding news articles is a crucial tool for multiple fields, such as media bias detection, identifying fake news, and making news recommendations. However, existing news embedding methods are not optimized to capture the latent context of news events. Most embedding methods rely on full-text information and neglect time-relevant embedding generation. In this paper, we propose a novel lightweight method that optimizes news embedding generation by focusing on entities and themes mentioned in articles and their historical connections to specific events. We suggest a method composed of three stages. First, we process and extract events, entities, and themes from the given news articles. Second, we generate periodic time embeddings for themes and entities by training time-separated GloVe models on current and historical data. Lastly, we concatenate the news embeddings generated by two distinct approaches: Smooth Inverse Frequency (SIF) for article-level vectors and Siamese Neural Networks for embeddings with nuanced event-related information. We leveraged over 850,000 news articles and 1,000,000 events from the GDELT project to test and evaluate our method. We conducted a comparative analysis of different news embedding generation methods for validation. Our experiments demonstrate that our approach can both improve and outperform state-of-the-art methods on shared event detection tasks.

Omri Porat, Michael Fire, Eran Ben-Elia

Dockless e-scooters, a key micromobility service, have emerged as eco-friendly and flexible urban transport alternatives. These services improve first and last-mile connectivity, reduce congestion and emissions, and complement public transport for short-distance travel. However, effective management of these services depends on accurate demand prediction, which is crucial for optimal fleet distribution and infrastructure planning. While previous studies have focused on analyzing spatial or temporal factors in isolation, this study introduces a framework that integrates spatial, temporal, and network dependencies for improved micromobility demand forecasting. This integration enhances accuracy while providing deeper insights into urban micromobility usage patterns. Our framework improves demand prediction accuracy by 27 to 49% over baseline models, demonstrating its effectiveness in capturing micromobility demand patterns. These findings support data-driven micromobility management, enabling optimized fleet distribution, cost reduction, and sustainable urban planning.

Danielle Kapon, Michael Fire, Shai Gordin

Cuneiform tablets, emerging in ancient Mesopotamia around the late fourth millennium BCE, represent one of humanity's earliest writing systems. Characterized by wedge-shaped marks on clay tablets, these artifacts provided insight into Mesopotamian civilization across various domains. Traditionally, the analysis and dating of these tablets rely on subjective assessment of shape and writing style, leading to uncertainties in pinpointing their exact temporal origins. Recent advances in digitization have revolutionized the study of cuneiform by enhancing accessibility and analytical capabilities. Our research uniquely focuses on the silhouette of tablets as significant indicators of their historical periods, diverging from most studies that concentrate on textual content. Utilizing an unprecedented dataset of over 94,000 images from the Cuneiform Digital Library Initiative collection, we apply deep learning methods to classify cuneiform tablets, covering over 3,000 years of history. By leveraging statistical, computational techniques, and generative modeling through Variational Auto-Encoders (VAEs), we achieve substantial advancements in the automatic classification of these ancient documents, focusing on the tablets' silhouettes as key predictors. Our classification approach begins with a Decision Tree using height-to-width ratios and culminates with a ResNet50 model, achieving a 61% macro F1-score for tablet silhouettes. Moreover, we introduce novel VAE-powered tools to enhance explainability and enable researchers to explore changes in tablet shapes across different eras and genres. This research contributes to document analysis and diplomatics by demonstrating the value of large-scale data analysis combined with statistical methods. These insights offer valuable tools for historians and epigraphists, enriching our understanding of cuneiform tablets and the cultures that produced them.

Shay Lapid, Dima Kagan, Michael Fire

Nowadays, detecting anomalous communities in networks is an essential task in research, as it helps discover insights into community-structured networks. Most of the existing methods leverage either information regarding attributes of vertices or the topological structure of communities. In this study, we introduce the Co-Membership-based Generic Anomalous Communities Detection Algorithm (referred as to CMMAC), a novel and generic method that utilizes the information of vertices co-membership in multiple communities. CMMAC is domain-free and almost unaffected by communities' sizes and densities. Specifically, we train a classifier to predict the probability of each vertex in a community being a member of the community. We then rank the communities by the aggregated membership probabilities of each community's vertices. The lowest-ranked communities are considered to be anomalous. Furthermore, we present an algorithm for generating a community-structured random network enabling the infusion of anomalous communities to facilitate research in the field. We utilized it to generate two datasets, composed of thousands of labeled anomaly-infused networks, and published them. We experimented extensively on thousands of simulated, and real-world networks, infused with artificial anomalies. CMMAC outperformed other existing methods in a range of settings. Additionally, we demonstrated that CMMAC can identify abnormal communities in real-world unlabeled networks in different domains, such as Reddit and Wikipedia.

Ran Ziv, Ilan Gronau, Michael Fire

Entity Matching is an essential part of all real-world systems that take in structured and unstructured data coming from different sources. Typically no common key is available for connecting records. Massive data cleaning and integration processes require completion before any data analytics, or further processing can be performed. Although record linkage is frequently regarded as a somewhat tedious but necessary step, it reveals valuable insights, supports data visualization, and guides further analytic approaches to the data. Here, we focus on organization entity matching. We introduce CompanyName2Vec, a novel algorithm to solve company entity matching (CEM) using a neural network model to learn company name semantics from a job ad corpus, without relying on any information on the matched company besides its name. Based on a real-world data, we show that CompanyName2Vec outperforms other evaluated methods and solves the CEM challenge with an average success rate of 89.3%.

Dima Kagan, Galit Fuhrmann Alpert, Michael Fire

The spread of the Red Palm Weevil has dramatically affected date growers, homeowners and governments, forcing them to deal with a constant threat to their palm trees. Early detection of palm tree infestation has been proven to be critical in order to allow treatment that may save trees from irreversible damage, and is most commonly performed by local physical access for individual tree monitoring. Here, we present a novel method for surveillance of Red Palm Weevil infested palm trees utilizing state-of-the-art deep learning algorithms, with aerial and street-level imagery data. To detect infested palm trees we analyzed over 100,000 aerial and street-images, mapping the location of palm trees in urban areas. Using this procedure, we discovered and verified infested palm trees at various locations.

Dima Kagan, Galit Fuhrmann Alpert, Michael Fire

The COVID-19 pandemic outbreak, with its related social distancing and shelter-in-place measures, has dramatically affected ways in which people communicate with each other, forcing people to find new ways to collaborate, study, celebrate special occasions, and meet with family and friends. One of the most popular solutions that have emerged is the use of video conferencing applications to replace face-to-face meetings with virtual meetings. This resulted in unprecedented growth in the number of video conferencing users. In this study, we explored privacy issues that may be at risk by attending virtual meetings. We extracted private information from collage images of meeting participants that are publicly posted on the Web. We used image processing, text recognition tools, as well as social network analysis to explore our web crawling curated dataset of over 15,700 collage images, and over 142,000 face images of meeting participants. We demonstrate that video conference users are facing prevalent security and privacy threats. Our results indicate that it is relatively easy to collect thousands of publicly available images of video conference meetings and extract personal information about the participants, including their face images, age, gender, usernames, and sometimes even full names. This type of extracted data can vastly and easily jeopardize people's security and privacy both in the online and real-world, affecting not only adults but also more vulnerable segments of society, such as young children and older adults. Finally, we show that cross-referencing facial image data with social network data may put participants at additional privacy risks they may not be aware of and that it is possible to identify users that appear in several video conference meetings, thus providing a potential to maliciously aggregate different sources of information about a target individual.

Aviad Elyashar, Rami Puzis, Michael Fire

Searching for information about a specific person is an online activity frequently performed by many users. In most cases, users are aided by queries containing a name and sending back to the web search engines for finding their will. Typically, Web search engines provide just a few accurate results associated with a name-containing query. Currently, most solutions for suggesting synonyms in online search are based on pattern matching and phonetic encoding, however very often, the performance of such solutions is less than optimal. In this paper, we propose SpokenName2Vec, a novel and generic approach which addresses the similar name suggestion problem by utilizing automated speech generation, and deep learning to produce spoken name embeddings. This sophisticated and innovative embeddings captures the way people pronounce names in any language and accent. Utilizing the name pronunciation can be helpful for both differentiating and detecting names that sound alike, but are written differently. The proposed approach was demonstrated on a large-scale dataset consisting of 250,000 forenames and evaluated using a machine learning classifier and 7,399 names with their verified synonyms. The performance of the proposed approach was found to be superior to 10 other algorithms evaluated in this study, including well used phonetic and string similarity algorithms, and two recently proposed algorithms. The results obtained suggest that the proposed approach could serve as a useful and valuable tool for solving the similar name suggestion problem.

Nadav Shalit, Michael Fire, Eran Ben-Elia

Public transport has become an essential part of urban existence with increased population densities and environmental awareness. Large quantities of data are currently generated, allowing for more robust methods to understand travel behavior by harvesting smart card usage. However, public transport datasets suffer from data integrity problems; boarding stop information may be missing due to imperfect acquirement processes or inadequate reporting. We developed a supervised machine learning method to impute missing boarding stops based on ordinal classification using GTFS timetable, smart card, and geospatial datasets. A new metric, Pareto Accuracy, is suggested to evaluate algorithms where classes have an ordinal nature. Results are based on a case study in the city of Beer Sheva, Israel, consisting of one month of smart card data. We show that our proposed method is robust to irregular travelers and significantly outperforms well-known imputation methods without the need to mine any additional datasets. Validation of data from another Israeli city using transfer learning shows the presented model is general and context-free. The implications for transportation planning and travel behavior research are further discussed.

Aviad Elyashar, Rami Puzis, Michael Fire

Searching for a person's name is a common online activity. However, Web search engines provide few accurate results to queries containing names. In contrast to a general word which has only one correct spelling, there are several legitimate spellings of a given name. Today, most techniques used to suggest synonyms in online search are based on pattern matching and phonetic encoding, however they often perform poorly. As a result, there is a need for an effective tool for improved synonym suggestion. In this paper, we propose a revolutionary approach for tackling the problem of synonym suggestion. Our novel algorithm, GRAFT, utilizes historical data collected from genealogy websites, along with network algorithms. GRAFT is a general algorithm that suggests synonyms using a graph based on names derived from digitized ancestral family trees. Synonyms are extracted from this graph, which is constructed using generic ordering functions that outperform other algorithms that suggest synonyms based on a single dimension, a factor that limits their performance. We evaluated GRAFT's performance on three ground truth datasets of forenames and surnames, including a large-scale online genealogy dataset with over 16 million profiles and more than 700,000 unique forenames and 500,000 surnames. We compared GRAFT's performance at suggesting synonyms to 10 other algorithms, including phonetic encoding, string similarity algorithms, and machine and deep learning algorithms. The results show GRAFT's superiority with respect to both forenames and surnames and demonstrate its use as a tool to improve synonym suggestion.

Michael Fire, Jonathan Schler

Online advertising is a huge, rapidly growing advertising market in today's world. One common form of online advertising is using image ads. A decision is made (often in real time) every time a user sees an ad, and the advertiser is eager to determine the best ad to display. Consequently, many algorithms have been developed that calculate the optimal ad to show to the current user at the present time. Typically, these algorithms focus on variations of the ad, optimizing among different properties such as background color, image size, or set of images. However, there is a more fundamental layer. Our study looks at new qualities of ads that can be determined before an ad is shown (rather than online optimization) and defines which ads are most likely to be successful. We present a set of novel algorithms that utilize deep-learning image processing, machine learning, and graph theory to investigate online advertising and to construct prediction models which can foresee an image ad's success. We evaluated our algorithms on a dataset with over 260,000 ad images, as well as a smaller dataset specifically related to the automotive industry, and we succeeded in constructing regression models for ad image click rate prediction. The obtained results emphasize the great potential of using deep-learning algorithms to effectively and efficiently analyze image ads and to create better and more innovative online ads. Moreover, the algorithms presented in this paper can help predict ad success and can be applied to analyze other large-scale image corpora.

Yuval Elovici, Michael Fire, Amir Herzberg et al.

Online Social Networks (OSNs) have rapidly become a prominent and widely used service, offering a wealth of personal and sensitive information with significant security and privacy implications. Hence, OSNs are also an important - and popular - subject for research. To perform research based on real-life evidence, however, researchers may need to access OSN data, such as texts and files uploaded by users and connections among users. This raises significant ethical problems. Currently, there are no clear ethical guidelines, and researchers may end up (unintentionally) performing ethically questionable research, sometimes even when more ethical research alternatives exist. For example, several studies have employed `fake identities` to collect data from OSNs, but fake identities may be used for attacks and are considered a security issue. Is it legitimate to use fake identities for studying OSNs or for collecting OSN data for research? We present a taxonomy of the ethical challenges facing researchers of OSNs and compare different approaches. We demonstrate how ethical considerations have been taken into account in previous studies that used fake identities. In addition, several possible approaches are offered to reduce or avoid ethical misconducts. We hope this work will stimulate the development and use of ethical practices and methods in the research of online social networks.