University of Northern British Columbia Institutional Repository

Search results

Search Results

Displaying (1 - 20 of 87)

Pages

Title: Sentence encoders for semantic textual similarity
Contributors: Aarthi Babu (author), Ian Hartley (thesis advisor), David Casperson (committee member), University of Northern British Columbia (Degree granting institution)
Abstract: The last decade has witnessed many accomplishments in the field of Natural Language Processing, especially in understanding the language semantics. Well-established machine learning models for generating word representation are available and has been proven useful. However, the existing techniques proposed for learning sentence level representations do not adequately capture the complexity of compositional semantics. Finding semantic similarity between sentences is a fundamental language understanding problem. In this project, we compare various machine learning models on their ability to capture the semantics of a sentence using Semantic Textual Similarity (STS) Task. We focus on models that exhibit state-of-the-art performance in Sem-Eval(2017) STS shared task. Also, we analyse the impact of models’ internal architectures on STS task performance. Out of all the models the we compared, Bi-LSTM RNN with max-pooling layer achieves the best performance in extracting a generic semantic representation and aids in better transfer learning when compared to hierarchical CNN.
Discipline: Computer Science

Title: Embedding programming languages: Prolog in Haskell
Contributors: Mehul Chandrakant Solanki (author), David Casperson (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis focuses on combining the two most important and wide spread declarative programming paradigms, functional and logic programming. The proposed approach aims at adding logic programming features which are native to Prolog onto Haskell. We develop extensions which replicate the target language by utilizing advanced features of the host language for an efficient implementation. The thesis aims to provide insights into merging two declarative languages namely, Haskell and Prolog by embedding the latter into the former and analyzing the results of doing so as the two languages have conflicting characteristics. The finished products will be something similar to a haskellised Prolog which has logic programming-like capabilities. --Leaf ii.
Discipline: Computer Science

Title: A basis for pronominal anaphora resolution using a model of working memory and long-term memory
Contributors: Clifford James Thompson (author), Charles Brown (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis presents a new theory of information modelling in natural language processing that attempts to resolve anaphoric references, while also addressing the problem of knowledge complexity. A modular model of semantic representation is introduced that addresses the deficiencies of existing representations, as well as the drawbacks associated with expanding these semantic representations. Rather than using a single semantic representation to model human knowledge and the knowledge within a sentence, the theory proposes a modular, multi-level model which is based around a semantic network. The behaviour of the model uses theories on the nature of working and long-term memory from cognitive psychology. Two methods of artificial neuron activation and decay were implemented - the ACT-R model and the Thompson model. Maximum success rates of 54.10% and 83.61% were achieved for The Three Brothers corpus, and maximum success rates of 56.00% and 86.67% were achieved for the Rumpelstiltskin corpus.
Discipline: Computer Science

Title: Lumping of atmospheric organic chemical species by machine learning.
Contributors: Pruthvi Polam (author), Margot Mandy (Thesis advisor), Charles Brown (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: No abstract available.
Discipline: Chemistry and Computer Science

Title: New watermarking methods for digital images.
Contributors: Md. Wahedul Islam (author), Saif Zahir (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: The phenomenal spread of the Internet places an enormous demand on content-ownership-validation. In this thesis, four new image-watermarking methods are presented. One method is based on discrete-wavelet-transformation (DWT) only while the rest are based on DWT and singular-value-decomposition (SVD) ensemble. The main target for this thesis is to reach a new blind-watermarking-method. Method IV presents such watermark using QR-codes. The use of QR-codes in watermarking is novel. The choice of such application is based on the fact that QR-Codes have errors self-correction-capability of 5% or higher which satisfies the nature of digital-image-processing. Results show that the proposed-methods introduced minimal distortion to the watermarked images as compared to other methods and are robust against JPEG, resizing and other attacks. Moreover, watermarking-method-II provides a solution to the detection of false watermark in the literature. Finally, method IV presents a new QR-code guided watermarking-approach that can be used as a steganography as well. --Leaf ii.
Discipline: Computer Science

Title: Improving Sonar Sensor Fidelity in a Robot Simulator.
Contributors: Allan Edward Kranz (author), Charles Brown (Thesis advisor), Liang Chen (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: It is slow and expensive to develop robot control systems using real robots. Simulation can provide the benefits of lowering the time and cost. In order to take advantage of the benefits of development in a simulator we need high fidelity representations of actual sensors. Sensors do not provide perfect data and simulations that use either perfect models or models that are too simple will not translate well into the real world. This research introduces a sensor model that overcomes some of the existing limitations in current simulations and provides a methodology for developing both new models and corresponding testing regimes. An actual sensor is used in realistic situations to create authentic models that more closely match the performance of the robot in the real world. A simple sonar sensor is tested against three generic obstacles and a realistic software simulation model of its capabilities is created. The Simbad robot simulator is modified to use this model, a testing regime is created to validate the results, and improved performance over the existing model is achieved. --Leaf iii.
Discipline: Computer Science

Title: Face recognition using convolutional macropixel comparison approach
Contributors: Yunke Li (author), Liang Chen (thesis advisor), University of Northern British Columbia (Degree granting institution), Jernej Polajnar (committee member), Jianbing Li (committee member)
Abstract: Convolutional Neural Network (CNN) is a widely used deep learning framework and is applied in the field of face recognition achieving outstanding results. Macropixel Comparison Approach is a shallow mathematical approach that recognizes face by comparing original pixel blocks of face images. In this thesis, we are inspired by ideas of the currently popular deep neural network framework and introduce two features into the mathematical approach: deep overlap and weighted filter. The aim is exploring if the idea of deep learning could benefit mathematical method which might extends the scope of face recognition research. Results from our experiments show that the new proposed approach achives markedly better recognition rates than the original macropixel method.
Discipline: Computer Science

Title: Individual versus group success in round robin tournament of an ability-based two-player game
Contributors: Jonathan Nyantakyi (author), Jernej Polajnar (thesis advisor), David Casperson (committee member), Christopher Opio (committee member), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis studies the development of cooperation and measures of success among self-interested agents in a defined ability-based two-player asymmetric game that is structured as a round-robin tournament. Our research is motivated by the notion that in many systems cooperative behaviour depends on some parameters that are usually not considered in existing research. These include: the balance between individual activity and interaction with others; the impact of agents' ability levels; and the need to maintain balance between individual and group performance. In this thesis, we examine all these issues by using a defined game-theoretic modelling and simulation framework. Our simulation experiments on six agent group compositions establish some patterns of how an agent's ability and strategy impact its individual and overall group performance. The results demonstrate that the design framework supports methodical comparative studies of strategy profiles with respect to specific individual and group performance measures.
Discipline: Computer Science

Title: A new highly efficient algorithm for lossless binary image compression.
Contributors: Lele Zhou (author), Saif Zahir (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: No abstract available.
Discipline: Computer Science

Title: A flexible simulaton framework for the study of deadlock resolution algorithms in multicore systems
Contributors: Dhruv Desai (author), Alex Aravind (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: Deadlock is a common phenomenon in software applications, yet it is ignored by most operating systems. Although the occurrence of a deadlocks in systems is not frequent, in some cases, the effects are drastic when deadlock occurs. The ongoing trend in processor technology indicates that future systems will have hundreds and thousands of cores. Due to this imminent trend in hardware development, the problem of deadlock has gained renewed attention in research. Deadlock handling techniques that are developed for earlier processors and distributed systems might not work well with multicore systems, due to their architectural differences. Hence, to maximize the utility of multicore systems, new programs have to be carefully designed and tested before they can be adopted for practical use. Many approaches have been developed to handle deadlock in multicore systems, but very little attention has been paid to comparing the performance of those approaches with respect to different performance parameters. To fulfil the above mentioned shortfalls, we need a flexible simulation testbed to study deadlock handling algorithms and to observe their performance differences in multicore systems. The development of such a framework is the main goal of this thesis. In the framework, we implemented a general a scenario, scenario for the Dining Philosopher's problem and scenario for the Banker's algorithm. In addition to these scenarios, we demonstrate the flexibility, soundness, and use of the proposed framework by simulating two different deadlock handling strategies "" deadlock avoidance (the Banker's algorithm) and deadlock detection (Dreadlocks). The deadlock detection is followed by deadlock recovery to resolve the detected deadlock. We also present result analysis for the different set of experiments performed on the implemented strategies. The proposed simulation testbed to study deadlocks in multicore systems is developed using Java. --Leaf i.
Discipline: Computer Science

Title: Multum in parvo: Toward a generic compression method for binary images.
Contributors: Arber Borici (author), Saif Zahir (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: Data compression is an active field of research as the requirements to efficiently store and retrieve data at minimum time and cost persist to date. Lossless or lossy compression of bi-level data, such as binary images, has an equally crucial factor of importance. In this work, we explore a generic, application-independent method for lossless binary image compression. The first component of the proposed algorithm is a predetermined fixed-size codebook comprising 8 x 8-bit blocks of binary images along with the corresponding codes of shorter lengths. The two variations of the codebook--Huffman codes and Arithmetic codes--have yielded considerable compression ratios for various binary images. In order to attain higher compression, we introduce a second component--the row-column reduction coding--which removes additional redundancy. The proposed method is tested on two major areas involving bi-level data. The first area of application consists of binary images. Empirical results suggest that our algorithm outperforms the standard JBIG2 by at least 5% on average. The second area involves images consisting of a predetermined number of discrete colors, such as digital maps and graphs. By separating such images into binary layers, we employed our algorithm and attained efficient compression down to 0.035 bits per pixel. --P.ii.
Discipline: Computer Science

Title: Face recognition using Dual Linear Regression based Classification
Contributors: Yuan Wang (author), Liang Chen (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This project discusses a face recognition algorithm that uses Dual Linear Regression based Classification (DLRC) and a voting approach. Each face image used is first converted into a cluster of images each image in the cluster is obtained by shifting the original image a few pixels. The similarity of a pair of face images can be measured by comparing the distance between the corresponding image clusters, which is calculated using DLRC approach. To further improve performance, each cluster of images, representing a single face image, is then partitioned into a union of clusters of sub images. DLRC is then used to measure similarities between corresponding sub-image clusters to provide temporary identity decisions a voting approach is applied to make final conclusions. We have carried out experiments on a benchmark dataset for face recognition. The result demonstrates that the proposed approach works best in certain simple situations, while its performance is also comparable to known algorithms in complicated situations. --Leaf ii.
Discipline: Computer Science

Title: An Interactive framework to develop and align business process models
Contributors: Dorob Wali Ahmad (author), Waqar Haque (thesis advisor), Alex Aravind (committee member), Balbinder Deo (committee member), University of Northern British Columbia (Degree granting institution)
Abstract: In the past few decades, the usage of Business Process Management (BPM) has enormously increased. Organizations are devoting resources towards development and use of BPM techniques and technologies to analyze, model, improve and implement business processes. The procedures currently used for collecting information to create business process models generally lead to misunderstanding or ambiguity between model and domain experts. We propose a framework to build business process models directly from users’ inputs captured through interactive web-forms. It also allows the users to align processes with strategic business objectives, critical success factors, and key performance indicators. Further, processes can be tagged with appropriate maturity level, types and tiers. The framework includes a dashboard with real-time reports which helps decision makers to monitor organization’s performance, make better decisions, and standardize/optimize processes across the organization. A comparison of the functionality available in different tools along with proposed framework is also presented.
Discipline: Computer Science

Title: Secure data communication over mobile devices in health networks.
Contributors: Ashish Sachdeva (author), Waqar Haque (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: The continuous developments in the field of mobile computing have made it possible to use mobile devices for healthcare applications. These devices can be used by healthcare providers to collect and share patients' medical data. However, with increasing adoption of mobile devices that carry confidential data, organizations need to secure the data from unauthorized users and mobile device theft. When unencrypted data is transmitted from one device to another it faces various security threats from malicious code, unsecure networks, unauthorized access, and data theft. The objective of this research is to develop a secure data sharing solution customized for healthcare environments, which would allow authorized users to securely access and share patients' data over mobile devices. We identify the vulnerable locations in mobile communication network that can possibly be exploited by unauthorized users or malicious code to access the confidential data, and develop an efficient security protocol that provides end to end data protection without compromising device's performance. To demonstrate the feasibility of our proposed data sharing architecture, a prototype customized for Point-of-Care-Testing (POCT) scenarios was built in collaboration with Northern Health, Prince George. Simulations were performed to analyze and validate our solution against the pre-defined requirement criteria. --P. ii.
Discipline: Computer Science

Title: A multiagent architecture for semantic query access to legacy relational databases.
Contributors: Mohammad Zubayer (author), Desenka Polajnar (Thesis advisor), Jernej Polajnar (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis proposes a novel approach to accessing information stored in legacy relational databases (RDB), based on Semantic Web and multiagent systems technologies. It introduces an architectural model of the Semantic Report Generation System (SRGS), designed to address the rising demand for flexible access to information in decision support systems. SRGS is composed of server Database Subsystems (DBS) and client User Subsystems (US). In a DBS, an agent interacts with the administrator to build a reference ontology from the RDB schema, which enables semantic queries without modifying the database. In a US, the decision-making user accesses the system through a simplified natural language interface, using customized extensions to the reference ontology that was imported from DBS an agent helps build the custom ontology, and facilitates query formulation and report generation. The proposed approach is illustrated by several scenarios that highlight the key behavioral aspects of accessing information and developing ontologies. --P. ii.
Discipline: Computer Science

Title: Prediction and preemptive control of network congestion in distributed real-time environment
Contributors: Ramandeep Dhanoa (author), Haque Waqar (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: Due to ever increasing demand for network capacity, the congestion problem is inflating. Congestion results in queuing within the network, packet loss and increased delays. It should be controlled to increase the system throughput and quality of service. The existing congestion control approaches such as source throttling and re-routing focus on controlling congestion after it has already happened. However, it is much more desirable to predict future congestion based on the current state and historical data, so that efficient controlling techniques can be applied to prevent congestion from happening in future. We have proposed a Neural Network Prediction-based routing (NNPR) protocol to predict as well as control the network traffic in distributed real time environment. A distributed real time transaction processing simulator (DRTTPS) has been used as the test-bed. For predictions, multi-step neural network model is developed in SPSS Modeler, which predicts congestion in future. ADAPA (Adaptive Decision and Predictive Analytics) scoring engine has been used for real-time scoring. An ADAPA wrapper calls the prediction model through web services and predicts the congestion in real-time. Once predicted results are obtained, messages are re-routed to prevent congestion. To compare our proposed work with existing techniques, two routing protocols are also implements "" Dijkstra's Shortest Path (DSP) and Routing Information Protocol (RIP). The main metric used to analyze the performance of our protocol is the percentage of transactions which complete before their deadline. The NNPR protocol is analyzed with various simulation runs having parameters both inside and outside the neural network input training range. Various parameters which can cause congestion were studied. These include bandwidth, worksize, latency, max active transactions, mean arrival time and update percentage. Through experimentation, it is observed that NNPR consistently outperforms DSP and RIP for all congestion loads. --Leaves [i]-ii.
Discipline: Computer Science

Title: An interaction protocol for bidirectional deliberation on direct help in agent teamwork.
Contributors: Mojtaba Malek Akhlagh (author), Jernej Polajnar (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis proposes a new interaction protocol for direct help in agent teamwork. It addresses design questions that may arise in practical systems development, and achieves higher teamwork performance impact than previous versions of the Mutual Assistance Protocol (MAP). Direct help, such as performing an action on teammate's behalf, is deliberated by team members as need arises, rather than imposed by team organization or centralized mechanisms. The deliberation can start with a request for help, or with an offer of help the two design principles have been embodied in two distinct versions of MAP. Based on their observed complementarity, we refine and combine them into a single protocol that leverages their individual advantages. Its novel features let an agent initiate help deliberation with request or offer, and also simultaneously provide and receive help. Simulation experiments demonstrate its team performance gains while varying the environment dynamism, agent resources, and communication costs. --Leaf i.
Discipline: Computer Science

Title: On-site tracking in wireless sensor networks.
Contributors: Baljeet Singh Malhotra (author), Alex Aravind (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: Tracking is one of the oldest practices that has evolved along with the advancement of technology. The task of tracking a moving target may have different objectives. Based on how these objectives are achieved, we classify the tracking problems into two broad categories: 'On-site tracking' and ' Off-site tracking'. The basic difference between these two approaches is the need for the physical presence of a mobile sink to track the target. This thesis deals with the On-site tracking problem in wireless sensor networks. First, we characterize the problem for a single target case and generalize it for the multiple targets case. Then, we propose a class of efficient algorithms to solve the problem. We also developed simulation software, OSTSim, for the performance study of the proposed algorithms. Theoretical bounds for the tracking time and the number of messages generated have been derived for our algorithms. An extensive simulation study shows that our algorithms are efficient.
Discipline: Computer Science

Title: A software system for agent-assisted ontology building
Contributors: Denish Mumbaiwala (author), Jernej Polajnar (Thesis advisor), Desanka Polajnar (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: This thesis investigates how one can design a team of intelligent software agents that helps its human partner develop a formal ontology from a relational database and enhance it with higher-level abstractions. The resulting efficiency of ontology development could facilitate the building of intelligent decision support systems that allow: high-level semantic queries on legacy relational databases autonomous implementation within a host organization and incremental deployment without affecting the underlying database or its conventional use. We introduce a set of design principles, formulate the prototype system requirements and architecture, elaborate agent roles and interactions, develop suitable design techniques, and test the approach through practical implementation of selected features. We endow each agent with model meta-ontology, which enables it to reason and communicate about ontology, and planning meta-ontology, which captures the role-specific know-how of the ontology building method. We also assess the maturity of development tools for a larger-scale implementation. --Leaf i.
Discipline: Computer Science

Title: Chess software and its impact on chess players
Contributors: Khaldoon Dhou (author), David Casperson (Thesis advisor), University of Northern British Columbia (Degree granting institution)
Abstract: Computer-aided chess is an important teaching method, as it allows a student to play under every condition possible, and regulates the speed of his/her development at an incremental pace, measured against actual players in the rated chess community. It is also relatively inexpensive, and pervasive, and allows players to match themselves against competitors from across the world. The learning process extends beyond games, as interactive software has shown it teaches several skills, such as opening, strategy, tactics, and chess-problem solving. Furthermore, current applications allow chess players to establish rankings via online chess tournaments, meet international grandmasters, and have access to training tools based on strategies from chess masters. Using 250 chess software packages, this research classifies them into distinct categories based mainly on the Gobet and Jansen's organization of the chess knowledge. This is followed by extensive discussion that analyzes these training tools, in order to identify the best training techniques available building on a research on human computer interaction, cognitive psychology, and chess theory. --P.ii.
Discipline: Computer Science