Numerical Analysis and Scientific Computing Commons^™

Approximate Inference In Collective Graphical Models, Daniel Sheldon, Tao Sun, Akshat Kumar, Thomas G. Dietterich

Research Collection School Of Computing and Information Systems

We study the problem of approximate inference in collective graphical models (CGMs), which were recently introduced to model the problem of learning and inference with noisy aggregate observations. We first analyze the complexity of inference in CGMs: unlike inference in conventional graphical models, exact inference in CGMs is NP-hard even for tree-structured models. We then develop a tractable convex approximation to the NP-hard MAP inference problem in CGMs, and show how to use MAP inference for approximate marginal inference within the EM framework. We demonstrate empirically that these approximation techniques can reduce the computational cost of inference by two orders …

A Direct Mining Approach To Efficient Constrained Graph Pattern Discovery, Feida Zhu, Zequn Zhang, Qiang Qu

Research Collection School Of Computing and Information Systems

Despite the wealth of research on frequent graph pattern mining, how to efficiently mine the complete set of those with constraints still poses a huge challenge to the existing algorithms mainly due to the inherent bottleneck in the mining paradigm. In essence, mining requests with explicitly-specified constraints cannot be handled in a way that is direct and precise. In this paper, we propose a direct mining framework to solve the problem and illustrate our ideas in the context of a particular type of constrained frequent patterns — the “skinny” patterns, which are graph patterns with a long backbone from which …

Mining User Relations From Online Discussions Using Sentiment Analysis And Probabilistic Matrix Factorization, Minghui Qiu, Liu Yang, Jing Jiang

Research Collection School Of Computing and Information Systems

Advances in sentiment analysis have enabled extraction of user relations implied in online textual exchanges such as forum posts. However, recent studies in this direction only consider direct relation extraction from text. As user interactions can be sparse in online discussions, we propose to apply collaborative filtering through probabilistic matrix factorization to generalize and improve the opinion matrices extracted from forum posts. Experiments with two tasks show that the learned latent factor representation can give good performance on a relation polarity prediction task and improve the performance of a subgroup detection task.

Your Love Is Public Now: Questioning The Use Of Personal Information In Authentication, Payas Gupta, Swapna Gottipati, Jing Jiang, Debin Gao

Research Collection School Of Computing and Information Systems

Most social networking platforms protect user's private information by limiting access to it to a small group of members, typically friends of the user, while allowing (virtually) everyone's access to the user's public data. In this paper, we exploit public data available on Facebook to infer users' undisclosed interests on their profile pages. In particular, we infer their undisclosed interests from the public data fetched using Graph APIs provided by Facebook. We demonstrate that simply liking a Facebook page does not corroborate that the user is interested in the page. Instead, we perform sentiment-oriented mining on various attributes of a …

It Is Not Just What We Say, But How We Say Them: Lda-Based Behavior-Topic Model, Minghui Qiu, Feida Zhu, Jing Jiang

Research Collection School Of Computing and Information Systems

Textual information exchanged among users on online social network platforms provides deep understanding into users' interest and behavioral patterns. However, unlike traditional text-dominant settings such as o ine publishing, one distinct feature for online social network is users' rich interactions with the textual content, which, unfortunately, has not yet been well incorporated in the existing topic modeling frameworks. In this paper, we propose an LDA-based behavior-topic model (B-LDA) which jointly models user topic interests and behavioral patterns. We focus the study of the model on online social network settings such as microblogs like Twitter where the textual content is relatively …

Directed Percolation And The Abstract Tile Assembly Model, Tyler Garrett Moore

Graduate Theses and Dissertations

Self-assembly is a process by which simple components build complex structures through local interactions. Directed percolation is a statistical physical model for describing competitive spreading processes on lattices. The author describes an algorithm which can transform a tile assembly system in the abstract Tile Assembly Model into a directed percolation problem, and then shows simulations of the aTAM which support this algorithm. The author also investigates two new constructs designed for Erik Winfree's abstract Tile Assembly Model called the NULL tile and temperature 1.5. These constructs aid the translation between self-assembly and directed percolation and may assist self-assembly researchers in …

Fans: Face Annotation By Searching Large-Scale Web Facial Images, Steven Hoi, Dayong Wang, I Yeu Cheng, Elmer Lin, Jianke Zhu, Ying He, Chunyan Miao

Research Collection School Of Computing and Information Systems

Auto face annotation is an important technique for many real-world applications, such as online photo album management, new video summarization, and so on. It aims to automatically detect human faces from a photo image and further name the faces with the corresponding human names. Recently, mining web facial images on the internet has emerged as a promising paradigm towards auto face annotation. In this paper, we present a demonstration system of search-based face annotation: FANS - Face ANnotation by Searching large-scale web facial images. Given a query facial image for annotation, we first retrieve a short list of the most …

Disclosing Climate Change Patterns Using An Adaptive Markov Chain Pattern Detection Method, Zhaoxia Wang, Gary Lee, Hoong Maeng Chan, Reuben Li, Xiuju Fu, Rick Goh, Pauline A. W. Poh Kim, Martin L. Hibberd, Hoong Chor Chin

Research Collection School Of Computing and Information Systems

This paper proposes an adaptive Markov chain pattern detection (AMCPD) method for disclosing the climate change patterns of Singapore through meteorological data mining. Meteorological variables, including daily mean temperature, mean dew point temperature, mean visibility, mean wind speed, maximum sustained wind speed, maximum temperature and minimum temperature are simultaneously considered for identifying climate change patterns in this study. The results depict various weather patterns from 1962 to 2011 in Singapore, based on the records of the Changi Meteorological Station. Different scenarios with varied cluster thresholds are employed for testing the sensitivity of the proposed method. The robustness of the proposed …

R-Energy For Evaluating Robustness Of Dynamic Networks, Ming Gao, Ee Peng Lim, David Lo

Research Collection School Of Computing and Information Systems

The robustness of a network is determined by how well its vertices are connected to one another so as to keep the network strong and sustainable. As the network evolves its robustness changes and may reveal events as well as periodic trend patterns that affect the interactions among users in the network. In this paper, we develop R-energy as a new measure of network robustness based on the spectral analysis of normalized Laplacian matrix. R-energy can cope with disconnected networks, and is efficient to compute with a time complexity of O (jV j + jEj) where V and E are …

Retweeting: An Act Of Viral Users, Susceptible Users, Or Viral Topics?, Tuan-Anh Hoang, Ee Peng Lim

Research Collection School Of Computing and Information Systems

When a user retweets, there are three behavioral factors that cause the actions. They are the topic virality, user virality and user susceptibility. Topic virality captures the degree to which a topic attracts retweets by users. For each topic, user virality and susceptibility refer to the likelihood that a user attracts retweets and performs retweeting respectively. To model a set of observed retweet data as a result of these three topic specific factors, we first represent the retweets as a three-dimensional tensor of the tweet authors, their followers, and the tweets themselves. We then propose the V 2S model, a …

Traditional Media Seen From Social Media, Jisun An, Daniele Quercia, Meeyoung Cha, Krishna Gummadi, Jon Crowcroft

Research Collection School Of Computing and Information Systems

With the advent of social media services, media outlets have started reaching audiences on social-networking sites. On Twitter, users actively follow a wide set of media sources, form interpersonal networks, and propagate interesting stories to their peers. These media subscription and interaction patterns, which had previously been hidden behind media corporations' databases, offer new opportunities to understand media supply and demand on a large scale. Through a map that connects 77 media outlets based on Twitter subscription patterns, we are able to answer a variety of questions: to what extent New York Times and the Wall Street Journal readers overlap? …

Image Processing Algorithms For Improving Planetary Exploration And Understanding, Ali Pouryazdanpanah

College of Engineering: Graduate Celebration Programs

• To design a fully automated tool-set that allows to detect and extract the sky region in planetary images.

• To develop the new method for rock segmentation in planetary stereo images.
• To develop the new method for shadow detection in planetary images

Mono-Sized Sphere Packing Algorithm Development Using Optimized Monte Carlo Technique, Karn Soontrapa, Yitung Chen

College of Engineering: Graduate Celebration Programs

In this research, fuel cell catalyst layer was developed using the optimized sphere packing algorithm. An optimization technique named adaptive random search technique (ARSET) was employed in this packing algorithm. The ARSET algorithm will generate the initial location of spheres and allow them to move in the random direction with the variable moving distance, randomly selected from the sampling range (a), based on the Lennard–Jones potential and Morse potential of the current and new configuration. The solid fraction values obtained from this developed algorithm are in the range of 0.610–0.624 while the actual processing time can significantly be reduced by …

Dynamic Label Propagation In Social Networks, Juan Du, Feida Zhu, Ee Peng Lim

Research Collection School Of Computing and Information Systems

Label propagation has been studied for many years, starting from a set of nodes with labels and then propagating to those without labels. In social networks, building complete user profiles like interests and affiliations contributes to the systems like link prediction, personalized feeding, etc. Since the labels for each user are mostly not filled, we often employ some people to label these users. And therefore, the cost of human labeling is high if the data set is large. To reduce the expense, we need to select the optimal data set for labeling, which produces the best propagation result. In this …

Measuring The Value Of Conceptual Modeling A Return On Investment (Roi) Analysis, Jeffrey Alan Brelsford

Computational Modeling & Simulation Engineering Theses & Dissertations

This thesis presents a method for calculating the return on investment (ROI) of conceptual modeling. Conceptual modeling as it relates to Modeling and Simulation (M&S) studies is a process that abstracts a model of a real or proposed system to be subsequently used for developing computer models. The conceptual modeling process is a critical constituent in an M&S project because it produces the central reference for building the computer model. Computer models require formal logic to carry out procedures, and consistent arguments in formal logic are crucial for the model to be useful. Not only does formal logic provide the …

Vistruclizer: A Structural Visualizer For Multi-Dimensional Social Networks, Bingtian Dai, Agus Trisnajaya Kwee, Ee Peng Lim

Research Collection School Of Computing and Information Systems

With the popularity of Web 2.0 sites, social networks today increasingly involve different kinds of relationships among different types of users in a single network. Such social networks are said to be multi-dimensional. Analyzing multi-dimensional networks is a challenging research task that requires intelligent visualization techniques. In this paper, we therefore propose a visual analytics tool called ViStruclizer to analyze structures embedded in a multi-dimensional social network. ViStruclizer incorporates structure analyzers that summarize social networks into both node clusters each representing a set of users, and edge clusters representing relationships between users in the node clusters. ViStruclizer supports user interactions …

Delayed Insertion And Rule Effect Moderation Of Domain Knowledge For Reinforcement Learning, Teck-Hou Teng, Ah-Hwee Tan

Research Collection School Of Computing and Information Systems

Though not a fundamental pre-requisite to efficient machine learning, insertion of domain knowledge into adaptive virtual agent is nonetheless known to improve learning efficiency and reduce model complexity. Conventionally, domain knowledge is inserted prior to learning. Despite being effective, such approach may not always be feasible. Firstly, the effect of domain knowledge is assumed and can be inaccurate. Also, domain knowledge may not be available prior to learning. In addition, the insertion of domain knowledge can frame learning and hamper the discovery of more effective knowledge. Therefore, this work advances the use of domain knowledge by proposing to delay the …

Roundtriprank: Graph-Based Proximity With Importance And Specificity, Yuan Fang, Kevin Chen-Chuan Chang, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Graph-based proximity has many applications with different ranking needs. However, most previous works only stress the sense of importance by finding "popular” results for a query. Often times important results are overly general without being well-tailored to the query, lacking a sense of specificity— which only emerges recently. Even then, the two senses are treated independently, and only combined empirically. In this paper, we generalize the well-studied importance-based random walk into a round trip and develop RoundTripRank, seamlessly integrating specificity and importance in one coherent process. We also recognize the need for a flexible trade-off between the two senses, and …

Finding The Optimal Social Trust Path For The Selection Of Trustworthy Service Providers In Complex Social Networks, Guanfeng Liu, Yan Wang, Mehmet A. Orgun, Ee Peng Lim

Research Collection School Of Computing and Information Systems

Online social networks have provided the infrastructure for a number of emerging applications in recent years, e.g., for the recommendation of service providers or the recommendation of files as services. In these applications, trust is one of the most important factors in decision making by a service consumer, requiring the evaluation of the trustworthiness of a service provider along the social trust paths from a service consumer to the service provider. However, there are usually many social trust paths between two participants who are unknown to one another. In addition, some social information, such as social relationships between participants and …

Csc Senior Project: Nlpstats, Michael Mease

Computer Science and Software Engineering

Natural Language Processing has recently increased in popularity. The field of authorship analysis, specifically, uses various characteristics of text quantified by markers. NLPStats serves as a tool designed to streamline marker extraction based on user needs. A flexible query system allows for custom marker requests, adjustment of result formatting, and preprocessing options. Furthermore, an efficiently designed structure ensures that users retrieve information quickly. As a whole, NLPStats enables anyone, regardless of NLP experience, to extract important information about the text of a document.