Digital Commons Network^™

Automatic Conversation Review For Intelligent Virtual Assistants, Ian R. Beaver

Computer Science ETDs

When reviewing the performance of Intelligent Virtual Assistants (IVAs), it is desirable to prioritize conversations involving misunderstood human inputs. These conversations uncover error in natural language understanding and help prioritize and expedite improvements to the IVA. As human reviewer time is valuable and manual analysis is time consuming, prioritizing the conversations where misunderstanding has likely occurred reduces costs and speeds improvement. A system for measuring the posthoc risk of missed intent associated with a single human input is presented. Numerous indicators of risk are explored and implemented. These indicators are combined using various means and evaluated on real world data. …

Al Planning Assistant For Scheduling Daily Activities, Priyanka Ahuja

Theses and Dissertations

Artificial conversational agents are software agents that can interact with humans in the way humans do. Siri Cortana, and Alexa are examples of intelligent agents that can help us with almost all the basic tasks. These agents are smart enough to do the basic tasks, but not as much when it comes to complex tasks, such as analyzing traffic data, reviewing scheduling conflicts, rescheduling meetings while resolving conflicts, and offering suggestions based upon data analyses (e.g. traffic patterns, weather, etc.) The actual potential of dialogue-based task agent potential remains untapped. The reason is the fact agents lack the ability to …

Hierarchical Learning Of Cross-Language Mappings Through Distributed Vector Representations For Code, Nghi D. Q. Bui, Lingxiao Jiang

Research Collection School Of Computing and Information Systems

Translating a program written in one programming language to another can be useful for software development tasks that need functionality implementations in different languages. Although past studies have considered this problem, they may be either specific to the language grammars, or specific to certain kinds of code elements (e.g., tokens, phrases, API uses). This paper proposes a new approach to automatically learn cross-language representations for various kinds of structural code elements that may be used for program translation. Our key idea is two folded: First, we normalize and enrich code token streams with additional structural and semantic information, and train …

Deep Code Comment Generation, Xing Hu, Ge Li, Xin Xia, David Lo, Zhi Jin

Research Collection School Of Computing and Information Systems

During software maintenance, code comments help developerscomprehend programs and reduce additional time spent on readingand navigating source code. Unfortunately, these comments areoften mismatched, missing or outdated in the software projects.Developers have to infer the functionality from the source code.This paper proposes a new approach named DeepCom to automatically generate code comments for Java methods. The generatedcomments aim to help developers understand the functionalityof Java methods. DeepCom applies Natural Language Processing(NLP) techniques to learn from a large code corpus and generatescomments from learned features. We use a deep neural networkthat analyzes structural information of Java methods for bettercomments generation. We conduct …

Entity Summarization Of Reviews And Micro-Reviews, Thanh Son Nguyen

Dissertations and Theses Collection (Open Access)

Along with the regular review content, there is a new type of user-generated content arising from the prevalence of mobile devices and social media, that is micro-review. Micro-reviews are bite-size reviews (usually under 200 char- acters), commonly posted on social media or check-in services, using a mobile device. They capture the immediate reaction of users, and they are rich in information, concise, and to the point. Both reviews and micro-reviews are useful for users to get to know the entity of interest, thus facilitating users in making their decision of purchasing or dining. However, the abundant number of both reviews …

Evaluating Sequence Discovery Systems In An Abstraction-Aware Manner, Eoin Rogers, Robert J. Ross, John D. Kelleher

Conference papers

Activity discovery is a challenging machine learning problem where we seek to uncover new or altered behavioural patterns in sensor data. In this paper we motivate and introduce a novel approach to evaluating activity discovery systems. Pre-annotated ground truths, often used to evaluate the performance of such systems on existing datasets, may exist at different levels of abstraction to the output of the output produced by the system. We propose a method for detecting and dealing with this situation, allowing for useful ground truth comparisons. This work has applications for activity discovery, and also for related fields. For example, it …

Parameterizing And Aggregating Activation Functions In Deep Neural Networks, Luke Benjamin Godfrey

Graduate Theses and Dissertations

The nonlinear activation functions applied by each neuron in a neural network are essential for making neural networks powerful representational models. If these are omitted, even deep neural networks reduce to simple linear regression due to the fact that a linear combination of linear combinations is still a linear combination. In much of the existing literature on neural networks, just one or two activation functions are selected for the entire network, even though the use of heterogenous activation functions has been shown to produce superior results in some cases. Even less often employed are activation functions that can adapt their …

Multimodal Depression Detection: An Investigation Of Features And Fusion Techniques For Automated Systems, Michelle Renee Morales

Dissertations, Theses, and Capstone Projects

Depression is a serious illness that affects a large portion of the world’s population. Given the large effect it has on society, it is evident that depression is a serious health issue. This thesis evaluates, at length, how technology may aid in assessing depression. We present an in-depth investigation of features and fusion techniques for depression detection systems. We also present OpenMM: a novel tool for multimodal feature extraction. Lastly, we present novel techniques for multimodal fusion. The contributions of this work add considerably to our knowledge of depression detection systems and have the potential to improve future systems by …

Adaptation And Augmentation: Towards Better Rescoring Strategies For Automatic Speech Recognition And Spoken Term Detection, Min Ma

Dissertations, Theses, and Capstone Projects

Selecting the best prediction from a set of candidates is an essential problem for many spoken language processing tasks, including automatic speech recognition (ASR) and spoken keyword spotting (KWS). Generally, the selection is determined by a confidence score assigned to each candidate. Calibrating these confidence scores (i.e., rescoring them) could make better selections and improve the system performance. This dissertation focuses on using tailored language models to rescore ASR hypotheses as well as keyword search results for ASR-based KWS.

This dissertation introduces three kinds of rescoring techniques: (1) Freezing most model parameters while fine-tuning the output layer in order to …

Evidence Aggregation For Answer Re-Ranking In Open-Domain Question Answering, Shuohang Wang, Mo Yu, Jing Jiang, Wei Zhang, Xiaoxiao Guo, Shiyu Chang, Zhiguo Wang, Tim Klinger, Gerald Tesauro, Murray Campbell

Research Collection School Of Computing and Information Systems

A popular recent approach to answering open-domain questions is to first search for question-related passages and then apply reading comprehension models to extract answers. Existing methods usually extract answers from single passages independently. But some questions require a combination of evidence from across different sources to answer correctly. In this paper, we propose two models which make use of multiple passages to generate their answers. Both use an answer-reranking approach which reorders the answer candidates generated by an existing state-of-the-art QA model. We propose two methods, namely, strength-based re-ranking and coverage-based re-ranking, to make use of the aggregated evidence from …

Proverbial Machine Translation: Translating Proverbs Between Spanish And English Using Phrased Based Statistical Machine Translation With The Grammatical Category Based Approach, Teneala Spencer

Honors Theses

The focus of the research presented in the paper is translating the non-literal interpretation of proverbs from Spanish to English without changing their intended meaning. Proverbs have limited variation in comparison to slang, poetry, and metaphors which tend to differ significantly within the context that they are used. Although there are many other approaches for machine translation (MT), the one most suitable for the research project is Phrased Based Statistical Machine translation used in conjunction with the Grammatical Category- Based approach.

Cognitive Virtual Admissions Counselor, Kumar Raja Guvindan Raju, Cory Adams, Raghuram Srinivas

SMU Data Science Review

Abstract. In this paper, we present a cognitive virtual admissions counselor for the Master of Science in Data Science program at Southern Methodist University. The virtual admissions counselor is a system capable of providing potential students accurate information at the time that they want to know it. After the evaluation of multiple technologies, Amazon’s LEX was selected to serve as the core technology for the virtual counselor chatbot. Student surveys were leveraged to collect and generate training data to deploy the natural language capability. The cognitive virtual admissions counselor platform is currently capable of providing an end-to-end conversational dialog to …

Mapping Ordinances And Tweets Using Smart City Characteristics To Aid Opinion Mining, Manish Puri, Xu Du, Aparna S. Varde, Gerard De Melo

Department of Computer Science Faculty Scholarship and Creative Works

This research focuses on mining ordinances (local laws) and public reactions to them expressed on social media. We place particular emphasis on ordinances and tweets relating to Smart City Characteristics (SCCs), since an important aim of our work is to assess how well a given region heads towards a Smart City. We rely on SCCs as a nexus between a seemingly infinite number of ordinances and tweets to be able to map them, and also to facilitate SCC-based opinion mining later for providing feedback to urban agencies based on public reactions. Common sense knowledge is harnessed in our approach to …

Automatic Analysis Of Musical Lyrics, Joanna Gormley

Honors Senior Capstone Projects

Is music getting less sophisticated over time? That is the question which this study aims to answer, with the goal of improving upon previous analysis done on the topic. The blog posts which inspired this project lacked accuracy and dimensionality. Realizing that a larger data set of songs would make a significant difference in the precision of our analysis, we set out to design a piece of software constructed with the capability to analyze several thousand songs. Mimicking previous works which analyzed sophistication of music, the software focuses on the lyrics of songs. Three metrics were used in order to …

Support Assurance-Based Software Development For Mission Critical Domains Using The Model Driven Architecture, Chung-Ling Lin

Dissertations

In the past decades, software development for mission critical applications has drawn great attention not only in various mission critical communities but also software engineering communities. One of the important reasons is that the failure of these systems can lead to some serious consequences such as huge financial loss and even loss of life. Therefore, software certification has become an important activity for mission critical applications in that software assurance for such a system should be certified. With the increasing complexity of a software system in mission critical sectors, certifiers have found hard time to understand how a software system …

Ontology Based Technical Skill Similarity, Yeshwanth Bashyam Balachander

Master's Projects

Online job boards have become a major platform for technical talent procurement and job search. These job portals have given rise to challenging matching and search problems. The core matching or search happens between technical skills of the job requirements and the candidate's profile or keywords. The extensive list of technical skills and its polyonymous nature makes it less effective to perform a direct keyword matching. This results in substandard job matching or search results which misses out a closely matching candidate on account of it not having the exact skills. It is important to use a semantic similarity measure …

Automated Lyrical Narrative Writing, Divya Singh

Master's Projects

Computational Creativity studies the potential of computers to act as autonomous creators and co-creators in addition to tools helping people. Creativity is evident in music, visual art, problem solving and languages. Significant work has been conducted in the area of linguistic creation mainly in the generation of stories, puns, rhymes, jokes, similes, and poetry. One of the major challenges of computational creativity is to generate lyrics that exhibit human-level creativity. On one hand, the lyrics generated should be meaningful and coherent, while on the other hand, they should satisfy poetry constraints such as rhyme scheme, rhyme type, and the number …

Subtopics In Yelp Reviews, Riya Suchdev

Master's Projects

Yelp is a review platform that connects people to local businesses. It is a very popular platform that helps customers decide which business to choose. It relies on crowd sourced plain text reviews. From the business’s description some facts can be determined, such as category and location. However, more detailed description can be extracted from the reviews. Discovering latent topics and subtopics in Yelp reviews, can help summarize the reviews to gain knowledge. For example, we can deduce that reviews related to the Restaurant category tend to emphasize on service, food, order etc. Additionally, one can deduce positive or negative …

Sentiment Analysis Using An Ensemble Of Feature Selection Algorithms, Manankumar Bhagat

Master's Projects

To determine the opinion of any person experiencing any services or buying any product, the usage of Sentiment Analysis, a continuous research in the field of text mining, is a common practice. It is a process of using computation to identify and categorize opinions expressed in a piece of text. Individuals post their opinion via reviews, tweets, comments or discussions which is our unstructured information. Sentiment analysis gives a general conclusion of audits which benefit clients, individuals or organizations for decision making. The primary point of this paper is to perform an ensemble approach on feature reduction methods identified with …

Does Journaling Encourage Healthier Choices? Analyzing Healthy Eating Behaviors Of Food Journalers, Palakorn Achananuparp, Ee Peng Lim, Vibhanshu Abhishek

Research Collection School Of Computing and Information Systems

Past research has shown the benefits of food journaling in promoting mindful eating and healthier food choices. However, the links between journaling and healthy eating have not been thoroughly examined. Beyond caloric restriction, do journalers consistently and sufficiently consume healthful diets? How different are their eating habits compared to those of average consumers who tend to be less conscious about health? In this study, we analyze the healthy eating behaviors of active food journalers using data from MyFitnessPal. Surprisingly, our findings show that food journalers do not eat as healthily as they should despite their proclivity to health eating and …

Using Sentiment Analysis And Pattern Matching To Signal User Review Abnormalities, Stefan S. Gloutnikov

Master's Projects

User opinions on websites like Amazon, Yelp, and TripAdvisor are a key input for consumers when figuring out what to purchase, or where and what to eat. This means that in order for such websites to provide a better service to their customers, they must guard against fake and targeted reviews. Detecting such users and reviews automatically is a very complex multi-step process, and there is no direct mechanism for solving the problem reliably. Multiple AI and Machine Learning algorithms are coupled together when examining user reviews in determining if a review is fake or not. In this project we …

Stock Price Prediction Using Deep Learning, Abhinav Tipirisetty

Master's Projects

Stock price prediction is one among the complex machine learning problems. It depends on a large number of factors which contribute to changes in the supply and demand. This paper presents the technical analysis of the various strategies proposed in the past, for predicting the price of a stock, and evaluation of a novel approach for the same. Stock prices are represented as time series data and neural networks are trained to learn the patterns from trends. Along with the numerical analysis of the stock trend, this research also considers the textual analysis of it by analyzing the public sentiment …

Detect Rumor And Stance Jointly By Neural Multi-Task Learning, Jing Ma, Wei Gao, Kam-Fai Wong

Research Collection School Of Computing and Information Systems

In recent years, an unhealthy phenomenon characterized as the massive spread of fake news or unverified information (i.e., rumors) has become increasingly a daunting issue in human society. The rumors commonly originate from social media outlets, primarily microblogging platforms, being viral afterwards by the wild, willful propagation via a large number of participants. It is observed that rumorous posts often trigger versatile, mostly controversial stances among participating users. Thus, determining the stances on the posts in question can be pertinent to the successful detection of rumors, and vice versa. Existing studies, however, mainly regard rumor detection and stance classification as …

Design And Implementation Of An Artificial Neural Network Controller For Quadrotor Flight In Confined Environment, Ahmed Mekky

Mechanical & Aerospace Engineering Theses & Dissertations

Quadrotors offer practical solutions for many applications, such as emergency rescue, surveillance, military operations, videography and many more. For this reason, they have recently attracted the attention of research and industry. Even though they have been intensively studied, quadrotors still suffer from some challenges that limit their use, such as trajectory measurement, attitude estimation, obstacle avoidance, safety precautions, and land cybersecurity. One major problem is flying in a confined environment, such as closed buildings and tunnels, where the aerodynamics around the quadrotor are affected by close proximity objects, which result in tracking performance deterioration, and sometimes instability. To address this …

Deep Learning For Chatbots, Vyas Ajay Bhagwat

Master's Projects

Natural Language Processing (NLP) requires modelling complex relationships between the semantics of the language. While traditional machine learning techniques are used for NLP, the models built for conversations, called chatbots, are unable to be truly generic. While chatbots have been made with traditional machine learning techniques, deep learning has allowed the complexities within NLP to be easier to model and can be leveraged to build a chatbot which has a real conversation with a human. In this project, we explore the problems and techniques used to build chatbots and where improvements can be made. We analyze different architectures to build …

Deep Learning Algorithm Recommender, Rajat Kabra

Master's Projects

Deep learning contains a set of algorithms that are based on the functioning of human brain i.e. neural networks. These algorithms require a lot of computation power and time along with complex setup to get good results. The project contains several artificial neural network implementation for a variety of tasks like data classification, image classification, natural language processing and more. The project contains an exploratory analysis of hyperparameters of deep learning algorithms in domain of deep learning applications to prove that it is possible to achieve a good accuracy with less resources.

Artificial Intelligence: An Analysis Of Alan Turing’S Role In The Conception And Development Of Intelligent Machinery, Erika L. Furtado

Selected Honors Theses

The purpose of this thesis is to follow the thread of Alan Turing’s ideas throughout his decades of research and analyze how his predictions have come to fruition over the years. Turing’s Computing Machinery and Intelligence is the paper in which the Turing Test is described as an alternative way to answer the question “can machines think?” (Turing 433). Since the development of Turing’s original paper, there has been a tremendous amount of advancement in the field of artificial intelligence. The field has made its way into art classification as well as the medical industry. The main concept researched in …

Perspective On And Re-Orientation Of Physical Proxies In Object-Focused Remote Collaboration, Martin Feick, Terrance Mok, Anthony Tang, Lora Oehlberg, Ehud Sharlin

Research Collection School Of Computing and Information Systems

Remote collaborators working together on physical objects have difficulty building a shared understanding of what each person is talking about. Conventional video chat systems are insufficient for many situations because they present a single view of the object in a flattened image. To understand how this limited perspective affects collaboration, we designed the Remote Manipulator (ReMa), which can reproduce orientation manipulations on a proxy object at a remote site. We conducted two studies with ReMa, with two main findings. First, a shared perspective is more effective and preferred compared to the opposing perspective offered by conventional video chat systems. Second, …

Eat & Tell: A Randomized Trial Of Random-Loss Incentive To Increase Dietary Self-Tracking Compliance, Palakorn Achananuparp, Ee Peng Lim, Vibhanshu Abhishek, Tianjiao Yun

Research Collection School Of Computing and Information Systems

A growing body of evidence has shown that incorporating behavioral economics principles into the design of financial incentive programs helps improve their cost-effectiveness, promote individuals' short-term engagement, and increase compliance in health behavior interventions. Yet, their effects on long-term engagement have not been fully examined. In study designs where repeated administration of incentives is required to ensure the regularity of behaviors, the effectiveness of subsequent incentives may decrease as a result of the law of diminishing marginal utility. In this paper, we introduce random-loss incentive-a new financial incentive based on loss aversion and unpredictability principles-to address the problem of individuals' …

Domain-Specific Cross-Language Relevant Question Retrieval, Bowen Xu, Zhenchang Xing, Xin Xia, David Lo, Shanping Li

Research Collection School Of Computing and Information Systems

Chinese developers often cannot effectively search questions in English, because they may have difficulties in translating technical words from Chinese to English and formulating proper English queries. For the purpose of helping Chinese developers take advantage of the rich knowledge base of Stack Overflow and simplify the question retrieval process, we propose an automated cross-language relevant question retrieval (CLRQR) system to retrieve relevant English questions for a given Chinese question. CLRQR first extracts essential information (both Chinese and English) from the title and description of the input Chinese question, then performs domain-specific translation of the essential Chinese information into English, …