Digital Commons Network^™

Leveraging Natural Language Processing Applications And Microblogging Platform For Increased Transparency In Crisis Areas, Ernesto Carrera-Ruvalcaba, Johnson Ekedum, Austin Hancock, Ben Brock

SMU Data Science Review

Through microblogging applications, such as Twitter, people actively document their lives even in times of natural disasters such as hurricanes and earthquakes. While first responders and crisis-teams are able to help people who call 911, or arrive at a designated shelter, there are vast amounts of information being exchanged online via Twitter that provide real-time, location-based alerts that are going unnoticed. To effectively use this information, the Tweets must be verified for authenticity and categorized to ensure that the proper authorities can be alerted. In this paper, we create a Crisis Message Corpus from geotagged Tweets occurring during 7 hurricanes …

A Hybrid Model To Detect Fake News, Indhumathi Gurunathan

Computer Science Graduate Projects and Theses

The wide availability of user-contributed content in the online social media facilitates aggregation of people around common interests, worldviews, and narratives. But over the years, internet being the source of information also becomes the source of misinformation. As people are generally awash in information, they can sometimes have difficulty discerning misinformation propagated on web platforms from truthful information. They may also lean heavily on information providers or social media platforms to curate information even though such providers do not commonly validate sources. In this project, we primarily focus was on political news and propose a hybrid model to detect misleading …

Twitter Bot Detection In The Context Of The 2018 Us Senate Elections, Wes Kendrick

Dartmouth College Undergraduate Theses

A growing percentage of public political communication takes place on social media sites such as Twitter, and not all of it is posted by humans. If citizens are to have the final say online, we must be able to detect and weed out bot accounts. The objective of this thesis is threefold: 1) expand the pool of Twitter election data available for analysis, 2) evaluate the bot detection performance of humans on a ground-truth dataset, and 3) learn what features humans associate with accounts that they believe to be bots. In this thesis, we build a large database of over …

Worcester Chamber Of Commerce: Recruiting Minority Business Owners, Ryan Dimaria, Alexander Hull, Xikun Lu, Haopeng Wang, Jiacheng Hou, Danning Zhao

School of Professional Studies

Our capstone project was to help the Worcester Regional Chamber of Commerce identify how to re-frame their marketing so it would be appealing to immigrant and minority owned businesses. Based on interviews and external research, our group was able to create a tangible and resourceful data set that provided justified recommendations and ideas on how the Chamber could make adjustments to their marketing plan to attract more businesses of this particular demographic in the city of Worcester. By implementing these recommendations, we believe the Chamber has the opportunity to create a more diverse group of Chamber members, add value to …

Frameannotator - A Web-Based Frame Semantic Annotation Tool, Sarbajit Roy

Computer Science and Engineering Theses - Archive

News headlines around the world are alarming. They aim to trigger emotional responses from the consumer. In an era where news and social media posts spread so quickly, it is often difficult to distinguish between what is real and what is false. Human fact-checking fails to cope with the growth of such unprecedented information, thus increasing the demand and various steps of automatic fact-checking to judge the veracity of claims. An automatic fact-checking system is a statistical model that can help to detect misinformation. The current state-of-the-art frame semantic parsers suffer from lack of a large annotated data-set and there …

Ultra-Context: Maximizing The Context For Better Image Caption Generation, Ankit Khare

Computer Science and Engineering Theses - Archive

Several combinations of visual and semantic attention have been geared towards developing better image captioning architectures. In this work we introduce a novel combination of word-level semantic context with image feature-level visual context, which provides a more holistic overall context for image caption generation. This approach does not require training any explicit network structure, using any external resource for training semantic attributes, or supervision during any training step. The proposed architecture addresses the significance of learning to find context at three levels to achieve a better trade-off as well as a balance between the two lines of attentiveness (word-level and …

Smart Home Audio Assistant, Xipeng Wang

All Computer Science and Engineering Research

This report introduces an audio processing algorithm. It provides a way to access smart devices using audio. Although there are many audio assistants already on the market, most of them will not be able to control the smart devices. Therefore, this new system presented in this report will provide a way to analysis the customer’s questions. Then the algorithm will be able to query smart device information, modify the schedule or provide the reason for some arrangement.

Commonsense Knowledge In Sentiment Analysis Of Ordinance Reactions For Smart Governance, Manish Puri

Theses, Dissertations and Culminating Projects

Smart Governance is an emerging research area which has attracted scientific as well as policy interests, and aims to improve collaboration between government and citizens, as well as other stakeholders. Our project aims to enable lawmakers to incorporate data driven decision making in enacting ordinances. Our first objective is to create a mechanism for mapping ordinances (local laws) and tweets to Smart City Characteristics (SCC). The use of SCC has allowed us to create a mapping between a huge number of ordinances and tweets, and the use of Commonsense Knowledge (CSK) has allowed us to utilize human judgment in mapping. …

Neural Multimodal Belief Tracker With Adaptive Attention For Dialogue Systems, Zheng Zhang, Lizi Liao, Minlie Huang, Xiaoyan Zhu, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

Multimodal dialogue systems are attracting increasing attention with a more natural and informative way for human-computer interaction. As one of its core components, the belief tracker estimates the user's goal at each step of the dialogue and provides a direct way to validate the ability of dialogue understanding. However, existing studies on belief trackers are largely limited to textual modality, which cannot be easily extended to capture the rich semantics in multimodal systems such as those with product images. For example, in fashion domain, the visual appearance of clothes play a crucial role in understanding the user's intention. In this …

Multimodal Review Generation For Recommender Systems, Quoc Tuan Truong, Hady W. Lauw

Research Collection School Of Computing and Information Systems

Key to recommender systems is learning user preferences, which are expressed through various modalities. In online reviews, for instance, this manifests in numerical rating, textual content, as well as visual images. In this work, we hypothesize that modelling these modalities jointly would result in a more holistic representation of a review towards more accurate recommendations. Therefore, we propose Multimodal Review Generation (MRG), a neural approach that simultaneously models a rating prediction component and a review text generation component. We hypothesize that the shared user and item representations would augment the rating prediction with richer information from review text, while sensitizing …

Processing Narratives By Means Of Action Languages, Craig Olson

Computer Science Theses, Dissertations, and Student Creative Activity

In this work we design a narrative understanding system Text2ALM that can be used in Question Answering domains. System Text2ALM utilizes an action language 𝒜ℒℳ to perform inferences on complex interactions of events described in narratives. The methodology that Text2ALM follows in its implementation was originally outlined by Yuliya Lierler, Daniela Inclezan, and Michael Gelfond in 2017 via a manual process, and this work serves as a proof of concept in a large-scale environment. Our system automates the conversion of a narrative to an 𝒜ℒℳ model containing facts about the narrative. We make use of the VerbNet lexicon that we …

Gogo: An Improved Algorithm To Measure The Semantic Similarity Between Gene Ontology Terms, Chenguang Zhao

Master's Theses

Measuring the semantic similarity between Gene Ontology (GO) terms is an essential step in functional bioinformatics research. We implemented a software named GOGO for calculating the semantic similarity between GO terms. GOGO has the advantages of both information-content-based and hybrid methods, such as Resnik’s and Wang’s methods. Moreover, GOGO is relatively fast and does not need to calculate information content (IC) from a large gene annotation corpus but still has the advantage of using IC. This is achieved by considering the number of children nodes in the GO directed acyclic graphs when calculating the semantic contribution of an ancestor node …

Emerging App Issue Identification From User Feedback: Experience On Wechat, Cuiyun Gao, Wujie Zheng, Yuetang Deng, David Lo, Jichuan Zeng, Michael R. Lyu, Irwin King

Research Collection School Of Computing and Information Systems

It is vital for popular mobile apps with large numbers of users to release updates with rich features while keeping stable user experience. Timely and accurately locating emerging app issues can greatly help developers to maintain and update apps. User feedback (i.e., user reviews) is a crucial channel between app developers and users, delivering a stream of information about bugs and features that concern users. Methods to identify emerging issues based on user feedback have been proposed in the literature, however, their applicability in industry has not been explored. We apply the recent method IDEA to WeChat, a popular messenger …

Deepjit: An End-To-End Deep Learning Framework For Just-In-Time Defect Prediction, Thong Hoang, Hoa Khanh Dam, Yasutaka Kamei, David Lo, Naoyasu Ubayashi

Research Collection School Of Computing and Information Systems

Software quality assurance efforts often focus on identifying defective code. To find likely defective code early, change-level defect prediction – aka. Just-In-Time (JIT) defect prediction – has been proposed. JIT defect prediction models identify likely defective changes and they are trained using machine learning techniques with the assumption that historical changes are similar to future ones. Most existing JIT defect prediction approaches make use of manually engineered features. Unlike those approaches, in this paper, we propose an end-to-end deep learning framework, named DeepJIT, that automatically extracts features from commit messages and code changes and use them to identify defects. Experiments …

Detect Rumors On Twitter By Promoting Information Campaigns With Generative Adversarial Learning, Jing Ma, Wei Gao, Kam-Fai Wong

Research Collection School Of Computing and Information Systems

Rumors can cause devastating consequences to individual and/or society. Analysis shows that widespread of rumors typically results from deliberately promoted information campaigns which aim to shape collective opinions on the concerned news events. In this paper, we attempt to fight such chaos with itself to make automatic rumor detection more robust and effective. Our idea is inspired by adversarial learning method originated from Generative Adversarial Networks (GAN). We propose a GAN-style approach, where a generator is designed to produce uncertain or conflicting voices, complicating the original conversational threads in order to pressurize the discriminator to learn stronger rumor indicative representations …

Concluding Remarks, Lei Meng, Ah-Hwee Tan, Donald C. Wunsch

Research Collection School Of Computing and Information Systems

This chapter summarizes the major contributions in this book and discusses their possible positions and requirements in some future scenarios. Section 8.1 follows the book structure to revisit the key contributions of this book in both theories and applications. The developed algorithms, such as the VA-ARTs for hyperparameter adaptation and the GHF-ART for multimedia representation and fusion, and the four applications, such as clustering and retrieving socially enriched multimedia data, are concentrated using one paragraph and three paragraphs, respectively. In Sect. 8.2, the roles of the proposed ART-embodied algorithms in social media clustering tasks are highlighted, and their possible evolutions …

Neural Multimodal Belief Tracker With Adaptive Attention For Dialogue Systems, Zheng Zhang, Lizi Liao, Minlie Huang, Xiaoyan Zhu, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

Multimodal dialogue systems are attracting increasing attention with a more natural and informative way for human-computer interaction. As one of its core components, the belief tracker estimates the user's goal at each step of the dialogue and provides a direct way to validate the ability of dialogue understanding. However, existing studies on belief trackers are largely limited to textual modality, which cannot be easily extended to capture the rich semantics in multimodal systems such as those with product images. For example, in fashion domain, the visual appearance of clothes play a crucial role in understanding the user's intention. In this …

Supervised Machine Learning Models For Fake News Detection, Gofaas Group, Andrea Lopez, Adelo Vieira, Zafar Ahsan, Farooq Saqib, Shirley Marinho

ICT

Fake news or the distribution of disinformation has become one of the most challenging issues in society. News and information are churned out across online websites and platforms in real-time, with little or no way for the viewing public to determine what is real or manufactured. But an awareness of what we are consuming online is becoming apparent and efforts are underway to explore how we separate fake content from genuine and truthful information.

The most challenging part of fake news is determining how to spot it. In technology, there are ways to help us do this. Supervised machine learning …

The Challenges Of Creating Engaging Content: Results From A Focus Group Study Of A Popular News Media Organization, Kholoud Khalil Aldous, Jisun An, Bernard J. Jansen

Research Collection School Of Computing and Information Systems

The process of content creation for distribution via social media platforms is not a trivial one for social media editors as the goal of creating both serious and engaging content is challenging, with no clear or differing guidelines or rules across and between platforms. For creators of serious content, such as news organizations, advertisers, or educational institutions, engagement has a deeper meaning beyond likes, shares, etc. that is aimed at the audience actually processing the underlying content associated with a social media post. In this research, we report findings from a group study that aimed to understand the process and …

Lab Practicum For Bias In Algorithms, Ameet Soni, Krista Karbowski Thomason

Digital Humanities Curricular Development

This is a course assignment to demonstrate potential biases encoded in algorithms (this can be linked more specifically to natural language processing, machine learning, or artificial intelligence) using the Word Embedding Association Test. In lab, students will work with programs that demonstrate the usefulness of word embedding algorithms in finding relationships between words. Then, students will use an implementation of the algorithm in "Semantics derived automatically from language corpora contain human-like biases" by Caliskan et al. to detect gender and racial bias encoded in word embeddings. The assignment has students design and run an experiment using the WEAT algorithm to …

Automatic Short Answer Grading Using Siamese Bidirectional Lstm Based Regression, Arya Prabhudesai, Nguyen Binh Duong Ta

Research Collection School Of Computing and Information Systems

Automatic student assessment plays an important role in education - it provides instant feedback to learners, and at the same time reduces tedious grading workload for instructors. In this paper, we investigate new machine learning techniques for automatic short answer grading (ASAG). The ASAG problem mainly involves assessing short, natural language responses to given questions automatically. While current research in the field has focused either on feature engineering or deep learning, we propose a new approach which combines the advantages of both. More specifically, we propose a Siamese Bidirectional LSTM Neural Network based Regressor in conjunction with handcrafted features for …

Question Answering With Textual Sequence Matching, Shuohang Wang

Dissertations and Theses Collection (Open Access)

Question answering (QA) is one of the most important applications in natural language processing. With the explosive text data from the Internet, intelligently getting answers of questions will help humans more efficiently collect useful information. My research in this thesis mainly focuses on solving question answering problem with textual sequence matching model which is to build vectorized representations for pairs of text sequences to enable better reasoning. And our thesis consists of three major parts.

In Part I, we propose two general models for building vectorized representations over a pair of sentences, which can be directly used to solve the …

Discursive Power In Contemporary Media Systems: A Comparative Framework, Andreas Jungherr, Oliver Posegga, Jisun An

Research Collection School Of Computing and Information Systems

Contemporary media systems are in transition. The constellation of organizations, groups, and individuals contributing information to national and international news flows has changed as a result of the digital transformation. The 'hybrid media system' has proven to be one of the most instructive concepts addressing this change. Its focus on the mutually dependent interconnections between various types of media organizations, actors, and publics has inspired prolific research. Yet the concept can tempt researchers to sidestep systematic analyses of information flows and actors’ differing degrees of influence by treating media systems as a black box. To enable large-scale, empirical comparative studies …

Planning For A Corpus Of Continuous Ratings Of Spoken Dialog Quality, Nigel G. Ward

Departmental Technical Reports (CS)

While many aspects of speech processing, including speech recognition and speech synthesis, have seen enormous advances over the past few years, advances in dialog have been more modest. This difference is largely attributable to the lack of resources that can support machine learning of dialog models and dialog phenomena. The research community accordingly needs a corpus of spoken dialogs with quality annotations every 100 milliseconds or so. We envisage a large and diverse collection: on the order of fifty hours of data, representing hundreds of speakers and many genres, with every instant labeled for interaction quality by one or more …

Self-Efficacy In Information Security: A Mixed Methods Study Of Deaf End-Users, Kyle Murbach

Masters Theses & Doctoral Dissertations

This explanatory sequential mixed methods study focuses on gaining an overall understanding of the potential variances in self-efficacy in information security and security practice behavior in the deaf population. Very little is understood about the deaf experience when engaging in security practices and their confidence levels in doing so. Due to the fastpaced nature of cyber security and its many facets, the human factor plays a crucial role in the success of cyber security. It is important to understand the potential implications of variances that may affect a deaf end-user’s security practice behavior to be able to provide more effective …

Conversational Agent: Developing A Model For Intelligent Agents With Transient Emotional States, Angie Dowdell

Theses and Dissertations

The demonstration of aggressive behaviors in the workplace can usually lead to negative professional consequences. However, combat-related professions may involve life and death decision making on an almost daily basis. This is one occupational area in which major duties may invoke or even require aggression. Today’s Army leaders face a different set of challenges than those faced by leaders of the past. Leaders of today must be trained to be both assertive and aggressive at the appropriate times while also displaying empathy. While this is not a new task for an Army leader, it is one that is approached differently …

Automatic Acquisition Of Annotated Training Corpora For Test-Code Generation, Magdalena Kacmajor, John D. Kelleher

Articles

Open software repositories make large amounts of source code publicly available. Potentially, this source code could be used as training data to develop new, machine learning-based programming tools. For many applications, however, raw code scraped from online repositories does not constitute an adequate training dataset. Building on the recent and rapid improvements in machine translation (MT), one possibly very interesting application is code generation from natural language descriptions. One of the bottlenecks in developing these MT-inspired systems is the acquisition of parallel text-code corpora required for training code-generative models. This paper addresses the problem of automatically synthetizing parallel text-code corpora …

Question Answering For Suicide Risk Assessment Using Reddit, Amanuel Alambo, Usha Lokala, Ugur Kursuncu, Krishnaprasad Thirunarayan, Amelia Gyrard, Randon S. Welton, Jyotishman Pathak, Amit P. Sheth

Kno.e.sis Publications

Mental Health America designed ten questionnaires that are used to determine the risk of mental disorders. They are also commonly used by Mental Health Professionals (MHPs) to assess suicidality. Specifically, the Columbia Suicide Severity Rating Scale (C-SSRS), a widely used suicide assessment questionnaire, helps MHPs determine the severity of suicide risk and offer an appropriate treatment. A major challenge in suicide treatment is the social stigma wherein the patient feels reluctance in discussing his/her conditions with an MHP, which leads to inaccurate assessment and treatment of patients. On the other hand, the same patient is comfortable freely discussing his/her mental …

Recommending New Features From Mobile App Descriptions, He Jiang, Jingxuan Zhang, Xiaochen Li, Zhilei Ren, David Lo, Xindong Wu, Zhongxuan Luo

Research Collection School Of Computing and Information Systems

The rapidly evolving mobile applications (apps) have brought great demand for developers to identify new features by inspecting the descriptions of similar apps and acquire missing features for their apps. Unfortunately, due to the huge number of apps, this manual process is time-consuming and unscalable. To help developers identify new features, we propose a new approach named SAFER. In this study, we first develop a tool to automatically extract features from app descriptions. Then, given an app, we leverage the topic model to identify its similar apps based on the extracted features and API names of apps. Finally, we design …

Comparelda: A Topic Model For Document Comparison, Maksim Tkachenko, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

A number of real-world applications require comparison of entities based on their textual representations. In this work, we develop a topic model supervised by pairwise comparisons of documents. Such a model seeks to yield topics that help to differentiate entities along some dimension of interest, which may vary from one application to another. While previous supervised topic models consider document labels in an independent and pointwise manner, our proposed Comparative Latent Dirichlet Allocation (CompareLDA) learns predictive topic distributions that comply with the pairwise comparison observations. To fit the model, we derive a maximum likelihood estimation method via augmented variational approximation …