The rapid development of machine learning has delivered new approaches, methods, and tools to multiple domains. I see potential for these developments, specifically natural language processing (NLP), to provide new insights, novel methods, and larger scale to social science research. However, novel NLP methods require substantial technical skills to implement. Some of the highest adoption of novel technical tools is in the area of social media analysis, where the volume of source material can overwhelm methods that rely on human capacity. My PhD dissertation aims to bridge the gap between NLP technologies and the unique needs of social science research by contributing to the development of an open-source NLP tool specifically tailored for social science researchers that reduces barriers to entry. The goal is to empower social science researchers by providing more opportunities to explore data in novel ways. This paper outlines the objectives, methodology, and expected outcomes of the proposed research study, which includes designing the development process, requirement analysis, prototyping an NLP tool, evaluating its usability and performance, and providing support for its integration into the research workflow.

Sustainability and large infrastructure projects are intricately linked. When considering large and disruptive infrastructure projects, proponents, regulators, and stakeholders consider a variety of information sources to understand how the project and its environmental changes will impact people living in the region. This is often done using small data, through interviews, town halls, and surveys. Big Data approaches have potential to augment these methods; for example, novel approaches natural language processing (NLP) on social media data can reveal human-nature interactions in impacted areas. Existing literature suggested that sentiment and frequency analyses can help us understand public attitudes. As a case study to illustrate this potential, we have examined public engagement with landscapes impacted by large hydroelectric development. We examined S6,122 geotagged image captions within 35 well-defined geographical locations associated with three dam projects in Canada: Mactaquac dam (New Brunswick), Oldman dam (Alberta), and Site C dam (British Columbia) on Instagram. The sentiment results of the study expressed a high positive attitude towards the three landscapes and frequency analysis exposed that the people are more likely to engage with the landscape in the context of fun, recreation, exercise activities and vacations. This can inform planning processes to ensure that impacts to these important interactions are examined and mitigated, achieving long-term environmental, social, and economic well-being.

A user experience checklist is not a substitute for an informed expert assessment or for user studies. But what does one do when there are too many apps and not enough experts? A guideline for novice assessors can aid in assessing large numbers of apps quickly, but there are few tools assessing the quality of user experience in information-presenting mobile applications (apps). The Mobile Application User Experience Checklist (MAUX-C) is an evidence-informed, user-friendly tool that measures the overall UX quality of apps. This paper describes the development and series of preliminary validation of MAUX-C. While first developed to be used as a standardized evaluation measure when conducting systematic reviews of information-presenting mobile apps, it could also be used as a reference during app development. This paper emphasizes that in addition to high-quality information, information-presenting apps need to provide a high-quality user experience to be taken seriously in the post-truth era.

Science is rapidly evolving, incorporating technology like autonomous vehicles, high-throughput scientific instruments, high-fidelity numerical models, and sensor networks, all generating data with increasing frequency, variety, and volume. Scientists committed to open science are interested in sharing this data, which requires research data infrastructure (RDI). The software underlying RDI is often created and/or deployed by people who have not received formal training in software engineering, or at organizations with primary mandates that do not include software development. Our understanding of software engineering as a field and practice does not universally translate to this software. As RDI software is pushed to handle larger data sets, and used to share data more widely, it is important to understand the maintainability, the resilience of the development community, and other indicators of long-term software project health. While there is a body of research on scientific software, and on free and open source software, it is not known if existing approaches to assessing these properties are effective for RDI software. In this exploratory study, we calculate one proxy measure for maintainability (code convention adherence) for a popular ocean data management system, and compare the results with four open source projects, and with the apparent experience of users as captured in public mailing lists and an issue tracker. The results advance our limited understanding of this type of software, and inform hypothesis generation and future research design.

Big Data analysis often relies on open data, integrating it with large private data sets, using it as ground truth information, or providing it as part of the input to large simulations. Data can be released openly by governments to achieve various objectives: transparency, informing citizen engagement, or supporting private enterprise, to name a few. To the latter objective, Big Data analytics algorithms rely on high-quality, timely access to various data sources, including open data. Examples include retail analytics drawing on open demographic data and weather forecast systems drawing on open weather and climate data. In this paper, we describe the rise of post-truth in society, and the risks this poses to the quality, integrity, and authenticity of open data. We also discuss approaches to identifying, assessing, and mitigating these risks, and suggest future steps to manage this data quality concern.

Ocean science is experiencing an explosion of data as researchers employ a widening variety of sensors, operating at higher fidelity and frequency, to inform our understanding of the global ocean. This is further complicated by the increasing integration of open science data from other disciplines to analyze complex systems, like climate change, animal migration, and sea/air interaction. This shift has been unplanned, chaotic, and emergent, and has placed the onus on researchers to stay current with best practices for managing, analyzing, and sharing data. Ocean scientists who do not have the technical skill to manage this data are turning to technologists, on the assumption they have the expertise required to help. To test this assumption, we examined an experiential learning program that placed technologists at ocean data centres in Canada, conducting interviews with students and employers to identify the competencies they believed were required to manage ocean data, which were missing in students' education up to that point, and which students gained during the work term placement.

As the spread of rumours has been increasing every day in online social networks (OSNs), it is important to analyze and understand this phenomenon. Damage caused by the spread of rumours is difficult to handle without a full understanding of the dynamics behind it. One of the central steps of understanding rumour spread is to analyze who spread rumours online, why, and how. In this research, we focus on the steps who and why by describing, implementing, and evaluating an approach that studies whether or not a group of users is actively involved in rumour discussions, and assesses rumour-spreading personality types in OSNs. We implement this general approach using Reddit data, and demonstrate its use by determining which users engage with a recurring rumour, and analyzing their comments using qualitative methods. We find that we can reliably classify users into one of three categories: (1) “Generally support a false rumour”, (2) “Generally refute a false rumour”, or (3) “Generally joke about a false rumour”. Combining text mining techniques, such as text classification, sentiment analysis, and social network analysis, we aim to identify and classify those rumour-spreading user categories automatically and provide a more holistic view of rumour spread in OSNs.

This paper introduces a new large-scale n-gram corpus that is created specifically from social media text. Two distinguishing characteristics of this corpus are its monthly temporal attribute and that it is created from 1.65 billion comments of user-generated text in Reddit. The usefulness of this corpus is exemplified and evaluated by a novel Topic-based Latent Semantic Analysis (TLSA) algorithm. The experimental results show that unsupervised TLSA outperforms all the state-of-the-art unsupervised and semi-supervised methods in SEMEVAL 2015: paraphrase and semantic similarity in Twitter tasks.

The objective of the Personalized Web Tasking (PWT) workshop series is to develop a community interested in the overlap between web services (as interoperable and easily composable components) and configurable personalizable applications. On the one hand, the core web-service community focuses primarily on enabling/supporting the processes around building distributed systems using services as the basic building blocks. On the other hand, user-centric communities have long been interested in enabling end users to customize their own experience using web resources, context-aware computing, and adaptive software, among others. PWT 2014, held during the IEEE World Congress on Services, focused on the overlap between the two areas, and brought together international researchers examining research relevant to this overlap.

Denial of Service (DoS) attacks overwhelm online services, preventing legitimate users from accessing a service, often with impact on revenue or consumer trust. Approaches exist to filter network-level attacks, but application-level attacks are harder to detect at the firewall. Filtering at this level can be computationally expensive and difficult to scale, while still producing false positives that block legitimate users. This article presents a model-based adaptive architecture and algorithm for detecting DoS attacks at the web application level and mitigating them. Using a performance model to predict the impact of arriving requests, a decision engine adaptively generates rules for filtering traffic and sending suspicious traffic for further review, where the end user is given the opportunity to demonstrate they are a legitimate user. If no legitimate user responds to the challenge, the request is dropped. Experiments performed on a scalable implementation demonstrate effective mitigation of attacks launched using a real-world DoS attack tool.

As the amount of data created and stored by organizations continues to increase, attention is turning to extracting knowledge from that raw data, including making some data available outside of the organization to enable crowd analytics. The adoption of the MapReduce paradigm has made processing Big Data more accessible, but is still limited to data that is currently available, often only within an organization. Fine-grained control over what information is shared outside an organization is difficult to achieve with Big Data, particularly in the MapReduce model. We introduce a novel approach to sharing that enables fine-grained control over what data is shared. Users submit analytics tasks that run on infrastructure near the actual data, reducing network bottlenecks. Organizations allow access to a logical version of their data created at runtime by filtering and transforming the actual data without creating storage-intensive stale copies, and resellers can further segment or augment this data to provide added value to analytics tasks. A loosely-coupled ecosystem driven by web services allows for discovery and sharing with a flexible, secure environment that limits the knowledge those running analytics need to have about the actual provider of the data. We describe a proof-of-concept implementation of the various components required to realize this ecosystem, and present a set of experiments to demonstrate feasibility, showing advantageous performance versus storage trade-offs.

To facilitate software development for multiple, federated cloud systems, abstraction layers have been introduced to mask the differences in the offerings, APIs, and terminology of various cloud providers. Such layers rely on a common ontology, which a) is difficult to create, and b) requires developers to understand both the common ontology and how various providers deviate from it. In this paper we propose and describe a structured query language for the cloud, Cloud SQL, along with a system and methodology for acquiring and organizing information from cloud providers and other entities in the cloud ecosystem such that it can be queried. It allows developers to run queries on data organized based on their semantic understanding of the cloud. Like the original SQL, we believe the use of a declarative query language will reduce development costs and make the multi-cloud accessible to a broader set of developers.

This paper presents a flexible deployment service for cloud computing. The service facilitates the specification and the execution of cloud deployment plans for applications. An application is described through a pattern, an abstract view that captures the logical view of the application and its mapping into cloud resources. The services instantiate the pattern in the cloud and allows for runtime updates of the deployment. The service is accessible through a RESTful interface. We identify the requirements for the service, describe its interfaces and show several case studies that capture the main features of the service.

The sharing of large and interesting Big Data in cloud environments can be achieved using data-as-a-service, where a provider offers data to interested users. In enhanced data-as-a-service, the data provider also supplies compute infrastructure, allowing users to run analytics tasks local to the data and reducing the (expensive and slow) transmission of data over networks. This paper describes a services-based ecosystem that allows providers to precisely share portions of their data with users, using a model where users submit MapReduce jobs that run on the provider's Hadoop infrastructure. Providers are given mechanisms to filter, segment, and/or transform data before it reaches the user's task. The ecosystem also allows for intermediaries who offer value-added filtrations, segmentations, or transformations of the data (for example, pre-filtering a dataset to only include high-income users). We describe the RESTful services required to enable this ecosystem, introduce a prototype to demonstrate the concept, and present experiments using this ecosystem to both provide and analyze different segments of a single large data set.

The service-oriented architecture style supports desirable qualities, including distributed, loosely coupled systems spanning organizational boundaries. Such systems and their configurations are challenging to understand, reason about, and test. Improved understanding of these systems will support activities such as autonomic runtime configuration, application deployment, and development/testing. Simulation is one way to understand and test service systems. This paper describes a literature survey of simulation frameworks for service-oriented systems, examining simulation software, systems, approaches, and frameworks used to simulate service-oriented systems. We identify a set of dimensions for describing the various approaches, considering their modeling methodology, their functionalities, their underlying infrastructure, and their evaluation. We then introduce the services-aware simulation framework (SASF), a simulation framework for predicting the behavior of service-oriented systems under different configurations and loads, and discuss the unique features that distinguish it from other systems in the literature. We demonstrate its use in simulating two service-oriented systems.

This paper introduces a cloud broker service (STRATOS) which facilitates the deployment and runtime management of cloud application topologies using cloud elements/services sourced on the fly from multiple providers, based on requirements specified in higher level objectives. Its implementation and use is evaluated in a set of experiments.

Descriptive information about available cloud services (i.e., metadata) is required in order to make good decisions about which cloud service provider(s) to utilize when deploying an application topology to the cloud. Presently, there are no uniform mechanisms for describing these services. Further, there is no unifying process that aggregates this metadata from the set of cloud providers and makes it available to a user in a programmatic fashion from a single location. This paper presents a methodology for and an implementation of a service-oriented application that provides relevant metadata information describing offered cloud services via a uniform RESTful web service. The data provided by this service is automatically acquired and mapped to a standard ontology. Community members can submit performance benchmarks using a metrics agent that submits metrics via a web service. Several example applications using this API to help users select resources are presented.

Maintainability is a desired property of software, and a variety of metrics have been proposed for measuring it, focusing on different notions of complexity and code readability. Many practices have been proposed to improve maintainability through code refactorings: improving the cohesion, simplification of interfaces, renamings to improve understandability. Code conventions are a body of advice on lexical and syntactic aspects of code, aiming to standardize low-level code design under the assumption that such a systematic approach will make code easier to read, understand, and maintain. We present the first stage in our examination of code-convention adherence practices as a proxy measurement for maintainability. Based on a preliminary survey of software engineers, we identify a set of coding conventions that most relate to maintainability. Then we devise a “convention adherence” metric, based on the number and severity of violations of a defined coding convention. Finally, we analyze several open-source projects according to this metric to better understand how consistent different teams are with respect to adopting and conforming to code conventions.

Text analysis is an important computational task, as unstructured data including text abound and can potentially provide interesting information and knowledge in a variety of areas. In our collaboration with Digital Humanists, we have started to examine the opportunities that the cloud offers to improving the response times of text-analysis tools so that users can comparatively analyze large text corpora across a variety of dimensions. To that end, we have started migrating existing text analysis tools to the cloud, beginning with TAPoR, the Text Analysis Portal for Research. In this paper, we discuss our experience redesigning and re-implementing four basic TAPoR operations on Hadoop and we report on the performance improvements enabled by the migration.

We describe a framework and methodology for managing the privacy policy of an enterprise, including creation (based on factors like legislation and consumer preferences), validation and verification, deployment and enforcement, and compliance testing for business processes and software. To evaluate this approach, one module of our framework (compliance testing) is implemented for an existing prominent electronic commerce software application. Our unique approach monitors the personal information sent and received by the software application and converts it to a standardized representation. At defined points in the electronic commerce workflow, the transmissions are compared to a set of privacy rules (extracted from a privacy policy) to ascertain compliance. Non-compliant transmissions of personal information are labeled `potential privacy infractions' and are reported. Though presently implemented for software testing, ultimately the methodology is intended to halt or alter a workflow to avoid privacy infractions.

Aspect oriented programming (AOP) supports the encapsulation of concerns cutting across the logical design of object-oriented software, by introducing aspects as a new unit of code. Code fragments, defined within aspects, such as advice or inter-type declarations, can be composed together with traditional system units, such as classes in OO code, using quantification expressions that specify where such fragments are to be injected and executed. The introduction of aspect-orientation in object-oriented software development raises two interesting comprehension questions. First, is the comprehension of AO software more difficult than traditional object-oriented software? Second, can aspects support comprehension tasks of object-oriented developers?