The rise in popularity of Internet of Things (IoT) devices has opened doors for privacy and security breaches in Cyber-Physical systems like smart homes, smart vehicles, and smart grids that affect our daily existence. IoT systems are also a source of big data that gets shared via the cloud. IoT systems in a smart home environment have sensitive access control issues since they are deployed in a personal space. The collected data can also be of a highly personal nature. Therefore, it is critical to building access control models that govern who, under what circumstances, can access which sensed data or actuate a physical system. Traditional access control mechanisms are not expressive enough to handle such complex access control needs, warranting the incorporation of new methodologies for privacy and security. In this paper, we propose the creation of the PALS system, that builds upon existing work in an attribute-based access control model, captures physical context collected from sensed data (attributes) and performs dynamic reasoning over these attributes and context-driven policies using Semantic Web technologies to execute access control decisions. Reasoning over user context, details of the information collected by the cloud service provider, and device type our mechanism generates as a consequent access control decisions. Our system’s access control decisions are supplemented by another sub-system that detects intrusions into smart home systems based on both network and behavioral data. The combined approach serves to determine indicators that a smart home system is under attack, as well as limit what data breach such attacks can achieve.

The post paper: Context Sensitive Access Control in Smart Home Environments appeared first on UMBC ebiquity.

Automating GDPR Compliance using Policy Integrated Blockchain

Data protection regulations, like GDPR, mandate security controls to secure personally identifiable information (PII) of the users which they share with service providers. With the volume of shared data reaching exascale proportions, it is challenging to ensure GDPR compliance in real-time. We propose a novel approach that integrates GDPR ontology with blockchain to facilitate real-time automated data compliance. Our framework ensures data operation is allowed only when validated by data privacy policies in compliance with privacy rules in GDPR. When a valid transaction takes place the PII data is automatically stored off-chain in a database. Our system, built using Semantic Web and Ethereum Blockchain, includes an access control system that enforces data privacy policy when data is shared with third parties.

The post paper: Automating GDPR Compliance using Policy Integrated Blockchain appeared first on UMBC ebiquity.

Temporal Understanding of Cybersecurity Threats

As cybersecurity-related threats continue to increase, understanding how the field is changing over time can give insight into combating new threats and understanding historical events. We show how to apply dynamic topic models to a set of cybersecurity documents to understand how the concepts found in them are changing over time. We correlate two different data sets, the first relates to specific exploits and the second relates to cybersecurity research. We use Wikipedia concepts to provide a basis for performing concept phrase extraction and show how using concepts to provide context improves the quality of the topic model. We represent the results of the dynamic topic model as a knowledge graph that could be used for inference or information discovery.

The post paper: Temporal Understanding of Cybersecurity Threats appeared first on UMBC ebiquity.

Reinforcement Quantum Annealing:
A Hybrid Quantum Learning Automata

We introduce the notion of reinforcement quantum annealing (RQA) scheme in which an intelligent agent searches in the space of Hamiltonians and interacts with a quantum annealer that plays the stochastic environment role of learning automata. At each iteration of RQA, after analyzing results (samples) from the previous iteration, the agent adjusts the penalty of unsatisfied constraints and re-casts the given problem to a new Ising Hamiltonian. As a proof-of-concept, we propose a novel approach for casting the problem of Boolean satisfiability (SAT) to Ising Hamiltonians and show how to apply the RQA for increasing the probability of finding the global optimum. Our experimental results on two different benchmark SAT problems (namely factoring pseudo-prime numbers and random SAT with phase transitions), using a D-Wave 2000Q quantum processor, demonstrated that RQA finds notably better solutions with fewer samples, compared to the best-known techniques in the realm of quantum annealing.

See also:

• publisher’s site;
• R. Ayanzadeh, Leveraging Artificial Intelligence to Advance Problem-Solving with Quantum Annealers, Phd dissertation, UMBC, 2020 .

The post Paper: Reinforcement Quantum Annealing: A Hybrid Quantum Learning Automata appeared first on UMBC ebiquity.

Reinforcement Quantum Annealing: A Quantum-Assisted Learning Automata Approach

We introduce the reinforcement quantum annealing (RQA) scheme in which an intelligent agent interacts with a quantum annealer that plays the stochastic environment role of learning automata and tries to iteratively find better Ising Hamiltonians for the given problem of interest. As a proof-of-concept, we propose a novel approach for reducing the NP-complete problem of Boolean satisfiability (SAT) to minimizing Ising Hamiltonians and show how to apply the RQA for increasing the probability of finding the global optimum. Our experimental results on two different benchmark SAT problems (namely factoring pseudo-prime numbers and random SAT with phase transitions), using a D-Wave 2000Q quantum processor, demonstrated that RQA finds notably better solutions with fewer samples, compared to state-of-the-art techniques in the realm of quantum annealing.

The post Reinforcement Quantum Annealing: A Quantum-Assisted Learning Automata Approach appeared first on UMBC ebiquity.

Ph.D. Dissertation Defense

Modeling and Extracting Information about Cybersecurity Events from Text

Taneeya Satyapanich

9:30-11:30 Monday, 18 November, 2019, ITE346?

People now rely on the Internet to carry out much of their daily activities such as banking, ordering food, and socializing with their family and friends. The technology facilitates our lives, but also comes with many problems, including cybercrimes, stolen data, and identity theft. With the large and increasing number of transactions done every day, the frequency of cybercrime events is also growing. Since the number of security-related events is too high for manual review and monitoring, we need to train machines to be able to detect and gather data about potential cyber threats. To support machines that can identify and understand threats, we need standard models to store the cybersecurity information and information extraction systems that can collect information to populate the models with data from text.

This dissertation makes two significant contributions. First, we defined rich cybersecurity event schema and annotated the news corpus following the schema. Our schema consists of event type definitions, semantic roles, and event arguments. Second, we present CASIE, a cybersecurity event extraction system. CASIE can detect cybersecurity events, identify event participants and their roles, including specifying realis values. It also groups the events, which are coreference.  CASIE produces output in easy to use format as a JSON object.

We believe that this dissertation will be useful for cybersecurity management in the future. It will quickly grasp cybersecurity event information out of the unstructured text and fill in the event frame. So we can compete with tons of cybersecurity events that happen every day.

Committee: Drs. Tim Finin (chair), Anupam Joshi, Tim Oates, Karuna Pande Joshi, Francis Ferraro

The post Defense: Taneeya Satyapanich, Modeling and Extracting Information about Cybersecurity Events from Text appeared first on UMBC ebiquity.

In my knowledge graph class yesterday we talked about the SPARQL query language and I illustrated it with DBpedia queries, including an example getting data about the movie Double Indemnity. I had brought a google assistant device and used it to compare its answers to those from DBpedia. When I asked the Google assistant “Who starred in the film Double Indemnity”, the first person it mentioned was Raymond Chandler. I knew this was wrong, since he was one of its screenwriters, not an actor, and shared an Academy Award for the screenplay. DBpedia’s data was correct and did not list Chandler as one of the actors.

I did not feel too bad about this — we shouldn’t expect perfect accuracy in these huge, general purpose knowledge graphs and at least Chandler played an important role in making the film.

After class I looked at the Wikidata page for Double Indemnity (Q478209) and saw that it did list Chandler as an actor. I take this as evidence that Google’s knowledge Graph got this incorrect fact from Wikidata, or perhaps from a precursor, Freebase.

The good news is that Wikidata had flagged the fact that Chandler (Q180377) was a cast member in Double Indemnity with a “potential Issue“. Clicking on this revealed that the issue was that Chandler was not known to have an occupation property that a “cast member” property (P161) expects, which includes twelve types, such as actor, opera singer, comedian, and ballet dancer. Wikidata lists chandler’s occupations as screenwriter, novelist, write and poet.

More good news is that the Wikidata fact had provenance information in the form of a reference stating that it came from CSFD (Q3561957), a “Czech and Slovak web project providing a movie database”. Following the link Wikidata provided led me eventually to the resource, which allowed my to search for and find its Double Indemnity entry. Indeed, it lists Raymond Chandler as one of the movie’s Hrají. All that was left to do was to ask for a translation, which confirmed that Hrají means “starring”.

Case closed? Well, not quite. What remains is fixing the problem.

The final good news is that it’s easy to edit or delete an incorrect fact in Wikidata. I plan to delete the incorrect fact in class next Monday. I’ll look into possible options to add an annotation in some way to ignore the incorrect ?SFD source for Chander being a cast member over the weekend.

Some possible bad news that public knowledge graphs like Wikidata might be exploited by unscrupulous groups or individuals in the future to promote false or biased information. Wikipedia is reasonably resilient to this, but the problem may be harder to manage for public knowledge graphs, which get much their data from other sources that could be manipulated.

The post Why does Google think Raymond Chandler starred in Double Indemnity? appeared first on UMBC ebiquity.

A semantically rich framework to enable real-time knowledge extraction from short length semi-structured documents

Lavana Elluri

10:30-11:30 Monday, 4 November 2019, ITE346

Knowledge is currently maintained as a large volume of unstructured text data in books, laws, regulations and policies, news and social media, academic and scientific reports, conversation and correspondence, etc. Most of these text documents are not often machine-processable. Hence it is hard to find relevant information from these texts quickly. Extracting and categorizing knowledge from the text of these numerous text stores requires significant manual effort and time. A critical open challenge that we propose to address is automated incremental text classification and identifying context from small documents. Our aim is to develop a semantically rich framework, including algorithms that will extract and classify the context of the text in real-time, to help enable users that update their policies regularly and organizations that are submitting proposals. We will use techniques from deep learning, semantic web, and natural language processing to build this framework. Our objectives include representing knowledge in cloud compliance / legal texts to create and populate a knowledge graph based on data protection regulations. Additionally, we will also correlate rules implemented in the referencing document with the rules in original policies to determine context similarity.

The post TALK: Real-time knowledge extraction from short semi-structured documents appeared first on UMBC ebiquity.

Automated Data Augmentation via Wikidata Relationships

Oyesh Singh, UMBC
10:30-11:30 Monday, 21 October 2019, ITE 346

With the increase in complexity of machine learning models, there is more need for data than ever. In order to fill this gap of annotated data-scarce situation, we look towards the ocean of free data present in Wikipedia and other WIkimedia resources. Wikipedia has an enormous amount of data in many languages along with the knowledge graph defined in Wikidata. In this presentation, I will explain how we utilized the Wikipedia/Wikidata data to boost the performance of BERT models for named entity recognition.

The post TALK: Automated Data Augmentation via Wikidata Relationships appeared first on UMBC ebiquity.

Knowledge Graph Fact Prediction via
Knowledge-Enriched Tensor Factorization

We present a family of novel methods for embedding knowledge graphs into real-valued tensors. These tensor-based embeddings capture the ordered relations that are typical in the knowledge graphs represented by semantic web languages like RDF. Unlike many previous models, our methods can easily use prior background knowledge provided by users or extracted automatically from existing knowledge graphs. In addition to providing more robust methods for knowledge graph embedding, we provide a provably-convergent, linear tensor factorization algorithm. We demonstrate the efficacy of our models for the task of predicting new facts across eight different knowledge graphs, achieving between 5% and 50% relative improvement over existing state-of-the-art knowledge graph embedding techniques. Our empirical evaluation shows that all of the tensor decomposition models perform well when the average degree of an entity in a graph is high, with constraint-based models doing better on graphs with a small number of highly similar relations and regularization-based models dominating for graphs with relations of varying degrees of similarity.

The post paper: Knowledge Graph Fact Prediction via Knowledge-Enriched Tensor Factorization appeared first on UMBC ebiquity.