Search

Electryo is a paper-based voting protocol that implements the Selene mechanism for individual verifiability. This short paper aims to provide the first formal model of Electryo, with security proofs for vote-privacy and individual verifiability. In general, voting protocols are complex constructs, involving advanced cryptographic primitives and strong security guarantees, posing a serious challenge when wanting to analyse and prove security with formal verification tools. Here we choose to use the \({{\,\mathrm{\textsc {Tamarin}}\,}}\) prover since it is one of the more advanced tools and is able to handle many of the primitives we encounter in the design and analysis of voting protocols.

When it comes to leaked credentials and credit card information, we observe the development and use of anti-fingerprinting browsers by malicious actors. These tools are carefully designed to evade detection, often by mimicking the browsing environment of the victim whose credentials were stolen. Even though these tools are popular in the underground markets, they have not received enough attention by researchers. In this paper, we report on the first evaluation of four underground, commercial, and research anti-fingerprinting browsers and highlight their high success rate in bypassing browser fingerprinting. Despite their success against well-known fingerprinting methods and libraries, we show that even slightest variation in the simulated fingerprint compared to the real ones can give away the presence of anti-fingerprinting tools. As a result, we provide techniques and fingerprint-based signatures that can be used to detect the current generation of anti-fingerprinting browsers.

This paper shows the automatic classification of research papers published in Scopus. Our classification is based on the research lines of the university. We apply the K-nearest neighbor classifier and linear discriminant analysis (LDA). Various stages were used from information gathering, creating the vocabulary, pre-processing and data training, and supervised classification in this work. The experiment involved 596 research articles published in SCOPUS from 2003–2017. The results show an overall accuracy of 88.44%.

Coreference resolution is essential for automatic text understanding to facilitate high-level information retrieval tasks such as text summarisation or question answering. Previous work indicates that the performance of state-of-the-art approaches (e.g. based on BERT) noticeably declines when applied to scientific papers. In this paper, we investigate the task of coreference resolution in research papers and subsequent knowledge graph population. We present the following contributions: (1) We annotate a corpus for coreference resolution that comprises 10 different scientific disciplines from Science, Technology, and Medicine (STM); (2) We propose transfer learning for automatic coreference resolution in research papers; (3) We analyse the impact of coreference resolution on knowledge graph (KG) population; (4) We release a research KG that is automatically populated from 55,485 papers in 10 STM domains. Comprehensive experiments show the usefulness of the proposed approach. Our transfer learning approach considerably outperforms state-of-the-art baselines on our corpus with an F1 score of 61.4 (+11.0), while the evaluation against a gold standard KG shows that coreference resolution improves the quality of the populated KG significantly with an F1 score of 63.5 (+21.8).

There have been a massive number of conferences and journals in computer science that create a lot of difficulties for scientists, especially for early-stage researchers, to find the most suitable venue for their scientific submission. In this paper, we present a novel approach for building a paper submission recommendation system by using two different types of embedding methods, GloVe and FastText, as well as Convolutional Neural Network (CNN) and LSTM to extract useful features for a paper submission recommendation model. We consider seven combinations of initial attributes from a given submission: title, abstract, keywords, title + keyword, title + abstract, keyword + abstract, and title + keyword + abstract. We measure these approaches’ performance on one dataset, presented by Wang et al., in terms of top K accuracy and compare our methods with the S2RSCS model, the state-of-the-art algorithm on this dataset. The experimental results show that CNN + FastText can outperform other approaches (CNN + GloVe, LSTM + GloVe, LSTM + FastText, S2RSCS) in term of top 1 accuracy for seven types of input data. Without using a list of keywords in the input data, CNN + GloVe/FastText can surpass other techniques. It has a bit lower performance than S2RSCS in terms of the top 3 and top 5 accuracies when using the keyword information. Finally, the combination of S2RSCS and CNN + FastText, namely S2CFT, can create a better model that bypasses all other methods by top K accuracy (K = 1,3,5,10).

In this paper we position Linear Time Temporal Logic (LTL), structural operational semantics (SOS), and a graphical generalization of BNF as central DSLs for program analysis and verification tasks in order to illustrate the impact of language to the mindset: (1) Specifying program analyses in LTL changes the classical algorithmic ‘HOW’ thinking into a property-oriented ‘WHAT’ thinking that allows one to logically combine analysis goals and eases proofs. (2) Playing with the original store component in SOS configurations allows one to elegantly realize variants of abstract program interpretations, and to align different aspects, like e.g., the symbolic values of variables and path conditions. (3) Specializing languages by refining their BNF-like meta models has the power to lift certain verification tasks from the program to the programming language level. We will illustrate the advantages of the change of mindset imposed by these three DSLs, as well as the fact that these advantages come at low price due to available adequate generator technology.

Game-based learning represents a promising alternative to teach computing in higher education. This paper presents “BDD Assemble!”, a paper-based game proposal for teaching Behavior Driven Development (BDD) competences. For this, the proposed game and the evaluation approach with software engineering students are described. As a result, a simple, interactive and colaborative game was provided, able to teach BDD concepts in a practical, competitive and fun way.

Technology is shaping our lives in a multitude of ways. This is fuelled by a technology infrastructure, both legacy and state of the art, composed of a heterogeneous group of hardware, software, services, and organisations. Such infrastructure faces a diverse range of challenges to its operations that include security, privacy, resilience, and quality of services. Among these, cybersecurity and privacy are taking the centre-stage, especially since the General Data Protection Regulation (GDPR) came into effect. Traditional security and privacy techniques are overstretched and adversarial actors have evolved to design exploitation techniques that circumvent protection. With the ever-increasing complexity of technology infrastructure, security and privacy-preservation specialists have started to look for adaptable and flexible protection methods that can evolve (potentially autonomously) as the adversarial actor changes its techniques. For this, Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) were put forward as saviours. In this paper, we look at the promises of AI, ML, and DL stated in academic and industrial literature and evaluate how realistic they are. We also put forward potential challenges a DL based security and privacy protection system has to overcome. Finally, we conclude the paper with a discussion on what steps the DL and the security and privacy-preservation community have to take to ensure that DL is not just going to be hype, but an opportunity to build a secure, reliable, and trusted technology infrastructure on which we can rely on for so much in our lives.

Co-citation analysis can be exploited as a bibliometric technique used for mining information on the relationships between scientific papers. Proposed methods rely, however, on co-citation counting techniques that slightly take the semantic aspect into consideration. The present study proposes a new technique based on the measure of Semantic Similarity (SS) between the titles of co-cited papers. Several computational measures rely on knowledge resources to quantify the semantic similarity, such as the WordNet «is a» taxonomy. Our proposal analyzes the SS between the titles of co-cited papers using word-based SS measures. Two major analytical experiments are performed: the first includes the benchmarks designed for testing word-based SS measures; the second exploits the dataset DBLP (DBLP: Digital Bibliography & Library Project.) citation network. As a result, we found the SS measures behave the same as human judgement for the lexical similarity and can be consequently used for the automatic assessment of similarity between co-cited papers. The analysis of highly repeated co-citations demonstrates that the different SS measures display almost similar behaviours, with slight differences due to the distribution of the provided SS values. Furthermore, we note a low percentage of similar referred papers into the co-citations.

DDoS attack detection using entropy-based features in network traffic has become a popular approach among researchers in the last five years. The use of traffic distribution features constructed using entropy measures has been proposed as a better approach to detect Distributed Denial of Service (DDoS) attacks compared to conventional volumetric methods, but it still lacks in the generality of detecting various intensity DDoS attacks accurately. In this paper, we focus on identifying effective entropy-based features to detect both high- and low-intensity DDoS attacks by exploring the effectiveness of entropy-based features in distinguishing the attack from normal traffic patterns. We hypothesise that using different entropy measures, window sizes, and entropy-based features may affect the accuracy of detecting DDoS attacks. This means that certain entropy measures, window sizes, and entropy-based features may reveal attack traffic amongst normal traffic better than the others. Our experimental results show that using Shannon, Tsallis and Zhou entropy measures can achieve a clearer distinction between DDoS attack traffic and normal traffic than Renyi entropy. In addition, the window size setting used in entropy construction has minimal influence in differentiating between DDoS attack traffic and normal traffic. The result of the effectiveness ranking shows that the commonly used features are less effective than other features extracted from traffic headers.

Patent terminology relation extraction is of great significance to the construction of patent Knowledge graph. In order to solve the problem of long-distance dependency in traditional depth learning, a new method of patent terminology relation extraction is proposed, which combines attention mechanism and bi-directional LSTM model and with keyword strategy. Category keyword features in each sentence obtained by the improved TextRank with the patent text information vectorization added. BiLSTM neural work and attention mechanism are employed to extract the temporal information and sentence-level global feature information. Moreover, pooling layer is added to obtain the local features of the text. Finally, we fuse the global features and local features, and output the final classification results through the softmax classifier. The addition of category keywords improves the distinction of categories. Substantial experimental results demonstrate that the proposed model outperform the state-of-art neural model in patent terminology relation extraction.

Understanding the current research trends, problems, and their innovative solutions remains a bottleneck due to the ever-increasing volume of scientific articles. In this paper, we propose NLPExplorer, a completely automatic portal for indexing, searching, and visualizing Natural Language Processing (NLP) research volume. NLPExplorer presents interesting insights from papers, authors, venues, and topics. In contrast to previous topic modelling based approaches, we manually curate five course-grained non-exclusive topical categories namely Linguistic Target (Syntax, Discourse, etc.), Tasks (Tagging, Summarization, etc.), Approaches (unsupervised, supervised, etc.), Languages (English, Chinese, etc.) and Dataset types (news, clinical notes, etc.). Some of the novel features include a list of young popular authors, popular URLs and datasets, list of topically diverse papers and recent popular papers. Also, it provides temporal statistics such as yearwise popularity of topics, datasets, and seminal papers. To facilitate future research and system development, we make all the processed dataset accessible through API calls. The current system is available at http://nlpexplorer.org.

The amount of data that produced is increased day after day especially data as a text, so with this massive production it would be difficult to analyze or extract information to discover the patterns from the unstructured text. Text mining is used for availing the massive amount of knowledge that is in the text and deriving high quality information from the text automatically. This Process would save effort and time. Text mining considered as a subset of data mining where data mining is more generic. This paper proposes a methodology of mining a text for a case study related to publication papers. Some of text mining approaches will be introduced for mining the publication papers using machine learning (ML) and natural language processing (NLP) techniques. Describing each phase as following: First phase is keywords extraction using natural language processing techniques, second phase named entity recognition and last phase is document classification. The last two phases are using the ML techniques. Then a case study is built to simulate the system phases, showing what is the input and the output in each phase.

Double-spending attacks on fast payments are one of the fatal architectural problems in Cryptocurrencies. Dmitrienko et al. proposed an offline fast payment scheme that relies on tamper-proof wallets produced by trustworthy manufacturers. With the wallets, the payee can immediately trust the transactions generated by the wallets without waiting for their registration to the blockchain. Secure coin-preloading to the wallet is important, while illegal coin-preloading can cause over/double-spending by the trusted wallets. For this, they proposed an interesting protocol that makes use of a fragment of the main blockchain to prove to the wallets the legitimacy of preloaded coins. One drawback is that, in proving that the fragment are from honest miners, their protocol requires a trusted online time-stamp server so that the wallets can verify the timestamps to see if the blocks in the fragment is mined with sufficiently large amount of computing resources. Otherwise, it sacrifices usability. In order to eliminate such an online trustee, in this paper we took the opposite approach that the payee (not the wallets) verifies the legitimacy of preloaded coins at the time of offline payment. As a consequence, our result shows that, with light-weight tamper-proof wallets, completely decentralized offline payment is possible without any modification to the existing Bitcoin network.

We revisit the 1997 PayWord credit-based micropayment scheme from Rivest and Shamir. We observe that smart contracts can be used to augment this system, apply to ‘claim or refund’ paradigm of cryptocurrencies to remove the counter-party risk inherent in PayWorld, and use a smart contract to ‘staple’ real value (in Ether) to payments in the system. Our implementation is more concise than any Ethereum payment channel we are aware of and the offline payments are very compact values (264 bits). It only uses hash functions and not digital signatures. EthWord becomes cheaper than standard Ethereum transfers when more than 16 payments between the same participants are made and appears to maintain its advantage for up to 1000+ transactions, at which point signature-based payments become cheapest. The main drawback of EthWord is the moderate gas price of using the system—despite dropping signatures, it is still priced out of the micropayments use-case. Like any payment channel, requires only two on-blockchain function calls to open and close the channel, while allowing the rest to be made off-blockchain.

We study blockchain-based integrity-protected smart contracts as an implementation mechanism for municipal government processes. To this end, we attempted a prototype implementation of such a process in collaboration with a Danish Municipality. We find that such an implementation is possible, despite the obvious confidentiality requirements, and that it does provide benefits: integrity guarantees, verifiability, direct collaboration and payments between the parties. These benefits come at the cost of latency, pr. transactions charges, immutability of errors, and a very concerning single point of failure the municipal government: losing blockchain private keys means losing control over municipal government casework, with no recourse. Our municipal government partner felt that altogether no immediately pressing problem was solved by the implementation, and that the latter risk clearly outweighed any benefits. We note that smart contract implementations of government processes needs to be immutable and outside of the government’s control when running; however, they also need to be updatable when laws change, and provide an “out” for the rare case when errors in the contract implementation result in unlawful behaviour. We propose these conflicting requirements as a foundational research challenge for blockchain to be applicable to governmental processes.

Blockchain technology enables a transaction to be handled in a decentralized fashion. In this position paper, we aim to introduce the background of blockchain technology, discuss one of its important component — smart contract, and present its recent applications in many fields such as cryptocurrency, financial services, risk management, and Internet of Things.

Scientific papers are important for scholars to track trends in specific research areas. With the increase in the number of scientific papers, it is difficult for scholars to read all the papers to extract emerging or noteworthy knowledge. Paper modeling can help scholars master the key information in scientific papers, and relation classification (RC) between entity pairs is a major approach to paper modeling. To the best of our knowledge, most of the state-of-the-art RC methods are using entire sentence’s context information as input. However, long sentences have too much noise information, which is useless for classification. In this paper, a flexible context is selected as the input information for convolution neural network (CNN), which greatly reduces the noise. Moreover, we find that entity type is another important feature for RC. Based on these findings, we construct a typical CNN architecture to learn features from raw texts automatically, and use a softmax function to classify the entity pairs. Our experiment on SemEval-2018 task 7 dataset yields a macro-F1 value of 83.91%, ranking first among all participants.

Unsupervised random-walk keyphrase extraction models mainly rely on global structural information of the word graph, with nodes representing candidate words and edges capturing the co-occurrence information between candidate words. However, integrating different types of useful information into the representation learning process to help better extract keyphrases is relatively unexplored. In this paper, we propose a random-walk method to extract keyphrases using word embeddings. Specifically, we first design a new word embedding learning model to integrate local context information of the word graph (i.e., the local word collocation patterns) with some crucial features of candidate words and edges. Then, a novel random-walk ranking model is designed to extract keyphrases by leveraging such word embeddings. Experimental results show that our approach outperforms 8 state-of-the-art unsupervised methods on two real datasets consistently for keyphrase extraction.

In the Android system security issue, the maliciousness of the applications is closely related to the permissions they applied. In this paper, a population-based model is proposed for detecting Android malicious application. Which is in the view of the current disadvantages of missing report, long detection period caused by features redundancy, and the instability of detection rate lead by unbalanced data of benign and malicious samples. Drawing on the idea of population in biology, each app was labeled by preprocessing. And adaptive feature vectors were automatically selected through the feature engineering. Thus the malicious application detection is carried out in the form of hybrid model voting. The experimental results show that feature engineering can remove a large amount of redundancy before classification. And the hybrid voting model can provide adaptive detection service for different populations.

Internet banking security is set to take a major step forward: On September 14, 2019, the Regulatory Technical Standards of the Revised Payment Service Directive (PSD2) are going to be effective within the European Union and the European Economic Area. This regulation makes two widely demanded transaction security properties mandatory: two-factor authentication, and the dynamic linking of the authentication code to the transaction’s beneficiary and amount (full transaction authentication). Even though the regulation is undoubtedly a positive development from a security perspective, it does not account for all the technical and human weak points involved in the transaction process. In this paper, we look at a series of attacks targeting online and mobile banking that are possible even in a post-PSD2 era. Despite the regulatory motivation of this work, the presented issues and suggestions to address them are likely to be universal for internet banking in general.

Trends show that cyber attacks targeting critical infrastructures are increasing, but security research for protecting such systems are challenging. There is a gap between the somewhat simplified models researchers at universities can sustain contra the complex systems at infrastructure owners that seldom can be used for direct research. There is also a lack of common datasets for research benchmarking. This paper presents a national experimental testbed for security research within supervisory control and data acquisition systems (SCADA), accessible for both research training and experiments. The virtualized testbed has been designed and implemented with both vendor experts and security researchers to balance the goals of realism with specific research needs. It includes a real SCADA product for energy management, a number of network zones, substation nodes, and a simulated power system. This environment enables creation of scenarios similar to real world utility scenarios, attack generation, development of defence mechanisms, and perhaps just as important: generating open datasets for comparative research evaluation.

Multi-view data that contains the data represented in many types of features has received much attention recently. The class of method utilising non-negative matrix factorization (NMF) and manifold learning to seek the meaningful latent structure of data has been popularly used for both traditional data and multi-view data. The NMF and manifold based multi-view clustering methods focus on dealing with the challenges of manifold learning and applying manifold learning on the NMF framework. This paper provides a comprehensive review of this important class of methods on multi-view data. We conduct an extensive experiments on several datasets and raise many open problems that can be dealt with in the future so a higher clustering performance can be achieved.

It is possible to relay signals between a contactless EMV card and a shop’s EMV reader and so make a fraudulent payment without the card-owner’s knowledge. Existing countermeasures rely on proximity checking: the reader will measure round trip times in message-exchanges, and will reject replies that take longer than expected (which suggests they have been relayed). However, it is the reader that would receive the illicit payment from any relayed transaction, so a rogue reader has little incentive to enforce the required checks. Furthermore, cases of malware targeting point-of-sales systems are common. We propose three novel proximity-checking protocols that use a trusted platform module (TPM) to ensure that the reader performs the time-measurements correctly. After running one of our proposed protocols, the bank can be sure that the card and reader were in close proximity, even if the reader tries to subvert the protocol. Our first protocol makes changes to the cards and readers, our second modifies the readers and the banking backend, and our third allows the detection of relay attacks, after they have happened, with only changes to the readers.

Monero is a privacy-centric cryptocurrency that makes payments untraceable by adding decoys to every real input spent in a transaction. Two studies from 2017 found methods to distinguish decoys from real inputs, which enabled traceability for a majority of transactions. Since then, a number protocol changes have been introduced, but their effectiveness has not yet been reassessed. Furthermore, little is known about traceability of Monero transactions across hard fork chains. We formalize a new method for tracing Monero transactions, which is based on analyzing currency hard forks. We use that method to perform a (passive) traceability analysis on data from the Monero, MoneroV and Monero Original blockchains and find that only a small amount of inputs are traceable. We then use the results to estimate the effectiveness of known heuristics for recent transactions and find that they do not significantly outperform random guessing. Our findings suggest that Monero is currently mostly immune to known passive attack vectors and resistant to tracking and tracing methods applied to other cryptocurrencies.

Machine learning is widely used in malware detection systems as a core component. However, machine learning algorithm is based on the assumption that the underlying malware concept is stable for training and testing. The assumption is vulnerable to well-crafted concept drift attacks, such as mimicry attacks, gradient descent attacks, poisoning attacks and so on. This paper proposes an ensemble learning system which combines vertical and horizontal correlation learning models. The significant diversity among vertical and horizontal correlation models increases the difficulty of concept drift attacks. And average p-value assessment is applied to fortify the system to be sensitive to hidden concept drift. The experiment results show that the hybrid system could actively recognize the concept drift among different Miuref variants.

The goal of anti-phishing techniques is to reduce the delivery rate of phishemails, and anti-phishing training aims to decrease the phishing click-through rates. This paper presents the X-Platform Phishing Attack, a deceptive phishing attack with an alarmingly high delivery and click-through rates, and highlights a subclass of phishing attacks that existing anti-phishing methods do not seem to be able to address. The main characteristic of this attack is that an attacker is able to embed a malicious link within a legitimate message generated by service providers (e.g., Github, Google, Amazon) and sends it using their infrastructure to his targets. This technique results in the bypassing of existing anti-phishing filters because it utilizes reputable service providers to generate seemingly legitimate emails. This also makes it highly likely for the targets of the attack to click on the phishing link as the email id of a legitimate provider is being used. An X-Platform Phishing attack can use email-based messaging and notification mechanisms such as friend requests, membership invitations, status updates, and customizable gift cards to embed and deliver phishing links to their targets. We have tested the delivery and click-through rates of this attack experimentally, based on a customized phishing email tunneled through GitHub’s pull-request mechanism. We observed that 100% of X-Platform Phishing emails passed the anti-phishing systems and were delivered to the inbox of the target subjects. All of the participants clicked on phishing messages, and in some cases, forwarded the message to other project collaborators who also clicked on the phishing links.

In recent years, following FAIR and open data principles, the number of available big data including biomedical data has been increased exponentially. In order to extract knowledge, these data should be curated, integrated, and semantically described. Accordingly, several semantic integration techniques have been developed; albeit effective, they may suffer from scalability in terms of different properties of big data. Even scaled-up approaches may be highly costly due to performing tasks of semantification, curation, and integration independently. To overcome these issues, we devise ConMap, a semantic integration approach which exploits knowledge encoded in ontologies to describe mapping rules in a way that performs all these tasks at the same time. The empirical evaluation of ConMap performed on different data sets shows that ConMap can significantly reduce the time required for knowledge graph creation by up to 70% of the time that is consumed following a traditional approach. Accordingly, the experimental results suggest that ConMap can be a semantic data integration solution that embody FAIR principles specifically in terms of interoperability.

The high availability and scalability of weakly-consistent system attracts system designers. Yet, writing correct application code for this type of systems is difficult; even how to specify the intended behavior of such systems is still an open question. There has not been established any standard method to specify the intended dynamic behavior of a weakly consistent system.

The rise of blockchain-based cryptocurrencies has led to an explosion of services using distributed ledgers as their underlying infrastructure. However, due to inherently single-service oriented blockchain protocols, such services can bloat the existing ledgers, fail to provide sufficient security, or completely forego the property of trustless auditability. Security concerns, trust restrictions, and scalability limits regarding the resource requirements of users hamper the sustainable development of loosely-coupled services on blockchains.

Predicting author’s attributes is useful for understanding implicit meanings of documents. The target problem of this paper is predicting author’s native language for each document. The authors of this paper used surface-level features of documents for the problem and tried to clarify the practical tendencies of the writing style as word occurrences. They conducted a classification of the abstracts written in English of approximately 85,000 scholarly papers written in English or in Japanese. As a result of the experiment, the accuracy of the binary classification was 0.97, and they found that a number of distinctive phrases used in the classification were related to typical writing styles of Japanese.

Thousands of academic papers are submitted at top venues each year. Manual audits are time-consuming and laborious. And the result may be influenced by human factors. This paper investigates a modularized and attention-based recurrent convolutional network model to represent academic paper and predict aspect scores. This model treats input text as module-document hierarchies, uses attention pooling CNN and LSTM to represent text, and outputs prediction with a linear layer. Empirical results on PeerRead data show that this model give the best performance among the baseline models.

Calculation of text similarity is an essential task for the text analysis and classification. It be can based, e.g., on Jaccard, cosine or other similar measures. Such measures consider the text as a bag-of-words and, therefore, lose some syntactic and semantic features of its sentences. This article presents a different measure based on the so-called artificial sentence pattern (ASP) method. This method has been developed to analyze texts in the Polish language which has very rich inflection. Therefore, ASP has utilized syntactic and semantic rules of the Polish language. Nevertheless, we argue that it admits extensions to other languages. As a result of the analysis, we have obtained several hypernodes which contain the most important words. Each hypernode corresponds to one of the examined documents, the latter being published papers from agriculture domain written in Polish. Experimental results obtained from that set of papers have been described and discussed. Those results have been visually illustrated using graphs of hypernodes and compared with Jaccard and cosine measures.

The citation similarity measurement task is defined as determining how similar the meanings of two citations are. This task play an significant role in Natural Language Processing applications, especially in academic plagiarism detection. Yet, computing citation similarity is not a trivial task, due to the incomplete and ambiguous information presented in academic papers, which makes necessity to leverage extra knowledge to understand it, as well as most similarity measures based on the syntactic features, and other based on the semantic part still has many drawbacks. In this paper, we propose a corpus-based approach using deep learning word embeddings to compute more effective citation similarity. Our study explores the previous works on text similarity, namely, string-based, knowledge-based and corpus-based. Then we define our new basis and experiment on a large dataset of scientific papers. The final results demonstrate that deep learning based approach can enhance the effectiveness of citation similarity.

Paper manufacturing is energy demanding and improved modelling of the pulp bleach process is the main non-invasive means of reducing energy costs. In this paper, time it takes to bleach paper pulp to desired brightness is examined. The model currently used is analysed and benchmarked against two machine learning models Random Forest and TreeBoost. Results suggests that the current model can be superseded by the machine learning models and it does not use the optimal compact subset of features. Despite the differences between the machine learning models, a feature ranking correlation has been observed for the new models. One novel, yet unused, feature that both machine learning models found to be important is the concentration of bleach agent.

A recent paper has proposed new ordering relations with uncertainty between the executions of tasks in acyclic process models. However, this approach cannot work for cyclic process models and those with silent transitions and non-free-choice constructs. In practice most non-trivial process models contain cycles and about 10 % to 20 % have also non-free-choice constructs. In this paper, we show how to overcome these problems by a refinement of the relations (i.e., extended refined ordering relations with uncertainty, ExRORU for short). All these relations can uniquely detect the behavioral differences between any pair of process models and can also be computed efficiently based on the complete prefix unfolding of a process model. Experiments on real-life and synthesized process models show that ExRORU is both effective and scalable.

The paper introduces the concept of \(\varPhi \)-data, data that is a proxy for some underlying data that offers advantages of data privacy and security while at the same time allowing particular data mining operations without requiring data owner participation once the proxy has been generated. The nature of the proxy representation is dependent on the nature of the desired data mining to be undertaken. Secure collaborative clustering is considered where the \(\varPhi \)-data is in the form of a Super Secure Chain Distance Matrices (SSCDM) encrypted using a proposed Multi-User Order Preserving Encryption (MUOPE) scheme. SSCDMs can be produced with respect to horizontal and vertical data partitioning. The DBSCAN clustering algorithm is adopted for illustrative and evaluation purposes. The results indicate that the proposed solution is efficient and produces comparable clustering configurations to those produced using an unencrypted, “standard”, algorithm; while maintaining data privacy and security.

The paper addresses the task of information extraction from scientific literature with machine learning methods. In particular, the tasks of definition and result extraction from scientific publications in Russian are considered. We note that annotation of scientific texts for creation of training dataset is very labor insensitive and expensive process. To tackle this problem, we propose methods and tools based on active learning. We describe and evaluate a novel adaptive density-weighted sampling (ADWeS) meta-strategy for active learning. The experiments demonstrate that active learning can be a very efficient technique for scientific text mining, and the proposed meta-strategy can be beneficial for corpus annotation with strongly skewed class distribution. We also investigate informative task-independent features for information extraction from scientific texts and present an openly available tool for corpus annotation, which is equipped with ADWeS and compatible with well-known sampling strategies.

We present a scenario where browsers’ network and computation capabilities are used by an attacker without the user’s knowledge. For this kind of abuse, an attacker needs to trigger JavaScript code on the browser, e.g. through an advertisement. However, unlike other Web attacks, e.g. cross-site scripting, the attack can be executed isolated from the Origin of the site visited by the user.

Clinical Guidelines are important knowledge resources for medical decision making. They provide clinical recommendations based on a collection of research findings with respect to a specific disease. Since, new findings are regularly published, CGs are also expected to be regularly updated. However, selecting and analysing medical publications require a huge human efforts, even when these publications are mostly regrouped and into repositories (e.g., MEDLINE database) and accessible via a search engine (e.g. PubMed). Automatically detecting those research findings from a medical search engine such as PubMed supports the guideline updating process. A simple search method is to select the medical terms that appear in the conclusions of the guideline to generate a query to search for new evidences. However, some challenges rise in this method: how to select the important terms, besides how to consider background knowledge that may be missing or not explicitly stated in those conclusions. In this paper we apply a knowledge model that formally describes elements such as actions and their effects to investigate (i) if it favors selecting the medical terms to compose queries and (ii) if a search enhanced with background knowledge can provide better result than other methods. This work explores a knowledge-driven approach for detecting new evidences relevant for the clinical guideline update process. Based on the outcomes of two experiments, we found that this approach can improve the recall by retrieving more relevant evidences than previous methods.

In this paper, we present \(\mathsf {FairScholar}\), a novel scientific paper recommendation system that aims at balancing both relevance and diversity while searching for research papers in response to keyword queries. Our system performs a vertex reinforced random-walk, a time heterogeneous random-walk on the citation graph of papers in order to factor in diversity while serving recommendations. To incorporate semantically similar items in the search results, it uses a query expansion step that finds similar keywords using community detection. An online demo of our search engine is available at http://www.cnergres.iitkgp.ac.in/FairScholar/.

This paper puts forward a vision of a universal ontology (UO) aiming at solving, or at least greatly alleviating, the semantic integration problem in the field of conceptual modeling and the understandability problem in the field of the semantic web. So far it has been assumed that the UO is not feasible in practice, but we think that it is time to revisit that assumption in the light of the current state-of-the-art. This paper aims to be a step in this direction. We try to make an initial proposal of a feasible UO. We present the scope of the UO, the kinds of its concepts, and the elements that could comprise the specification of each concept. We propose a modular structure for the UO consisting of four levels. We argue that the UO needs a complete set of concept composition operators, and we sketch three of them. We also tackle a few issues related to the feasibility of the UO, which we think that they could be surmountable. Finally, we discuss the desirability of the UO, and we explain why we conjecture that there are already organizations that have the knowledge and resources needed to develop it, and that might have an interest in its development in the near future.

We present Paper2vec, a novel neural network embedding based approach for creating scientific paper representations which make use of both textual and graph-based information. An academic citation network can be viewed as a graph where individual nodes contain rich textual information. With the current trend of open-access to most scientific literature, we presume that this full text of a scientific article contain vital source of information which aids in various recommendation and prediction tasks concerning this domain. To this end, we propose an approach, Paper2vec, which comprises of information from both the modalities and results in a rich representation for scientific papers. Over the recent past representation learning techniques have been studied extensively using neural networks. However, they are modeled independently for text and graph data. Paper2vec leverages recent research in the broader field of unsupervised feature learning from both graphs and text documents. We demonstrate the efficacy of our representations on three real world academic datasets in two tasks - node classification and link prediction where Paper2vec is able to outperform state-of-the-art by a considerable margin.

Successfully employed in the industry, hardware isolation environment enhances the access control of traditional operating systems and requires more rigorous analysis. This paper first applies the B method to the access control mechanism formalization and proposes an extensible formal model, which not only specifies the access control mechanism with process state transition in Linux, but also introduces the hardware isolation description. Consistent with program implementations, the B specifications can be animated and verified. The proposed B model constructs a mathematical framework for the security analysis, providing a theoretical support for mechanism enhancements. All the model components are type checked by Atelier B, with 547 proof obligations automatically generated. The current rate of model proof is 79%. The experimental results by ProB show that there is no invariant violation or deadlock. In conclusion, this paper presents a feasible solution for access control mechanism formalization and verification in the embedded system design. The access control model can be further extended and refined, with its specifications transformed into executable codes after proved.

This work presents a new, scalable solution to the problem of extracting citation contexts: the textual fragments surrounding citation references. These citation contexts can be used to navigate digital libraries of research papers to help users in deciding what to read. We have developed a prototype system which can retrieve, on-demand, citation contexts from the full text of over 15 million research articles in the Mendeley catalog for a given reference research paper. The evaluation results show that our citation extraction system provides additional functionality over existing tools, has two orders of magnitude faster runtime performance, while providing a 9% improvement in F-measure over the current state-of-the-art.

Modern cryptographic schemes have been focusing on protecting attacks from computational bounded adversaries. The various cryptographic primitives are designed concretely following some randomization design strategies, so that one of the goals is to make it hard for the attacker to distinguish between the real ciphers and the randomly distributed ones. Recently, Google Brain team proposed the idea to build cryptographic scheme automatically based on the neural network, and they claim that the scheme can defeat neural network adversaries. While it is a whole new direction, the security of the underlined scheme is remained unknown. In this paper, we investigate their basic statistical behavior from traditional cryptography’s point of view and extend their original scheme to discuss how the encryption protocol behave under a much more stronger adversary.

Understanding the possible associations between two entities from a query is a hard problem. For instance, querying “coffee” and “cancer” even in a curated Digital Library is a challenge to the retrieval system that struggles to figure out the intention of the query. Maybe the user wants a consensus of what it is known? But how many different associations exist? How to find them all? Herein we introduce an approach to diversify the results retrieved from such queries aiming at re-ranking the result list. Our re-ranking models specifically one fundamental aspect of scientific papers: claims. Claims are the sentences that scientists use to report findings. In particular, we study claims that express associations between entities in the medical domain. More specifically, we focus on queries that involve two entities in which one of the entities has some effect on a disease. Thus, we work on a corpus obtained by querying PubMed to empirically assess our proposed solution. Moreover, we promote the idea of claims as an explicit key aspect to consider diversification in the result set of a query. We show the potential of our approach to ease the process of discovering representative associations between entities. Our approach relies on a representation of claims using neural embedding of word vectors and implements an algorithm to perform the re-ranking of the result set of a query. We empirically show the potential of our approach.

This paper presents an attribute-based encryption (ABE) construction for monotone span programs achieving the shortest known ciphertext size under well-studied static complexity assumptions. Our ABE construction is built in composite order bilinear group setting and involves only 2 group elements in the ciphertexts. For proving selective security of the proposed ABE scheme under the Subgroup Decision assumptions, the most standard static assumptions in composite order bilinear group setting, we apply the extended version of the elegant Deja Q framework, which was originally proposed as a general technique for reducing the q-type complexity assumptions to their static counter parts. Our work thus demonstrates the power of this framework in overcoming the need of q-type assumptions, which are vulnerable to serious practical attacks, for deriving security of highly expressive ABE systems with compact parameters. We further introduce the concept of online-offline multi-input functional encryption (OO-MIFE), which is a crucial advancement towards realizing this highly promising but computationally intensive cryptographic primitive in resource bounded and power constrained devices. We also instantiate our notion of OO-MIFE by constructing such a scheme for the multi-input analog of the inner product functionality, which has a wide range of application in practice. Our OO-MIFE scheme for multi-input inner products is built in asymmetric bilinear groups of prime order and is proven selectively secure under the well-studied k -Linear assumption.

Cyber threat intelligence sharing is an imperative process to survive current and future attacks. The received information may protect stakeholders from being attacked by utilizing the course of action to remedy on-site vulnerabilities. Automating this process has shown to be challenging because several processes have to be synchronized and orchestrated to achieve the goal of automated information sharing. Organizations are inundated with threat information generated on site and received through crowd sourcing. This work presents a novel component for automated sharing, i.e. the content relevance filter.

Transport Layer Security (TLS) is the main standard designed for secure connections over the Internet. Security of TLS connections against active Man-in-the-Middle attacks relies on correctly validating public-key certificates during TLS handshake authentication. Although Certificate Transparency (CT) and further improved CT system—IKP mitigated the certificate authentication issues from the perspective of monitoring CA misbehavior, less attentions have been paid to consider the misbehavior of domain in using certificates during TLS handshake authentication. One misusing case is that domains refuse to use the certificates in Certificate Transparency Log for their own profits, the other is that a malicious domain impersonates the real one to deceive clients. In order to defend against domain’s misbehaviors in using certificates, we propose ETDA system based on IKP and CT aiming to enhance the security of TLS protocol from a novel perspective. ETDA is a blockchain-based system enforcing the automatic punishments in response to domain misbehavior and compensations to the client during TLS handshake authentication. The decentralized nature and incentives mechanism of ETDA provide an effective approach to prevent domains from sending invalid certificates to clients. We implement this system through Ethereum platform and Game Theory, which proved to be both technically and economically feasible.