Open Access. Powered by Scholars. Published by Universities.®

Digital Commons Network™

Open Access. Powered by Scholars. Published by Universities.®

Physical Sciences and Mathematics

Institution
Keyword
Publication Year
Publication
Publication Type
File Type

Articles 3841 - 3870 of 6325

Full-Text Articles in Entire DC Network

A Recurrent Neural Network Architecture For Biomedical Event Trigger Classification, Jeevith Bopaiah Jan 2018

A Recurrent Neural Network Architecture For Biomedical Event Trigger Classification, Jeevith Bopaiah

Theses and Dissertations--Computer Science

A “biomedical event” is a broad term used to describe the roles and interactions between entities (such as proteins, genes and cells) in a biological system. The task of biomedical event extraction aims at identifying and extracting these events from unstructured texts. An important component in the early stage of the task is biomedical trigger classification which involves identifying and classifying words/phrases that indicate an event. In this thesis, we present our work on biomedical trigger classification developed using the multi-level event extraction dataset. We restrict the scope of our classification to 19 biomedical event types grouped under four broad …


Predicting Perceived Age: Both Language Ability And Appearance Are Important, Sarah Plane, Ariel Marvasti, Tyler Egan, Casey Kennington Jan 2018

Predicting Perceived Age: Both Language Ability And Appearance Are Important, Sarah Plane, Ariel Marvasti, Tyler Egan, Casey Kennington

Computer Science Faculty Publications and Presentations

When interacting with robots in a situated spoken dialogue setting, human dialogue partners tend to assign anthropomorphic and social characteristics to those robots. In this paper, we explore the age and educational level that human dialogue partners assign to three different robotic systems, including an un-embodied spoken dialogue system. We found that how a robot speaks is as important to human perceptions as the way the robot looks. Using the data from our experiment, we derived prosodic, emotional, and linguistic features from the participants to train and evaluate a classifier that predicts perceived intelligence, age, and education level.


New Trends In Second Language Learning And Teaching Through The Lens Of Ict, Networked Learning, And Artificial Intelligence, Jaya Kannan, Pilar Munday Jan 2018

New Trends In Second Language Learning And Teaching Through The Lens Of Ict, Networked Learning, And Artificial Intelligence, Jaya Kannan, Pilar Munday

Languages Faculty Publications

In the last few decades, Information and Communications Technology (ICT) applications have been shaping the field of Computer Assisted Language Learning (CALL). Mobile Assisted Language Learning (MALL) paved the way for ubiquitous learning. The advent of new technologies in the early 21st century also added a social dimension to ICT that allowed for Networked Learning (NL). Given that language learning is fundamentally a socio-cultural experience, networked learning capabilities have provided the potential for language learning in community settings. This has revitalized the earlier frameworks provided by CALL. NL has empowered language learners today to connect globally, to access Open Educational …


Deep Neural Networks For Multi-Label Text Classification: Application To Coding Electronic Medical Records, Anthony Rios Jan 2018

Deep Neural Networks For Multi-Label Text Classification: Application To Coding Electronic Medical Records, Anthony Rios

Theses and Dissertations--Computer Science

Coding Electronic Medical Records (EMRs) with diagnosis and procedure codes is an essential task for billing, secondary data analyses, and monitoring health trends. Both speed and accuracy of coding are critical. While coding errors could lead to more patient-side financial burden and misinterpretation of a patient’s well-being, timely coding is also needed to avoid backlogs and additional costs for the healthcare facility. Therefore, it is necessary to develop automated diagnosis and procedure code recommendation methods that can be used by professional medical coders.

The main difficulty with developing automated EMR coding methods is the nature of the label space. The …


Assesing Completeness Of Solvency And Financial Condition Reports Through The Use Of Machine Learning And Text Classification, Ruairí Nugent Jan 2018

Assesing Completeness Of Solvency And Financial Condition Reports Through The Use Of Machine Learning And Text Classification, Ruairí Nugent

Dissertations

Text mining is a method for extracting useful information from unstructured data through the identification and exploration of large amounts of text. It is a valuable support tool for organisations. It enables a greater understanding and identification of relevant business insights from text. Critically it identifies connections between information within texts that would otherwise go unnoticed. Its application is prevalent in areas such as marketing and political science however, until recently it has been largely overlooked within economics. Central banks are beginning to investigate the benefits of machine learning, sentiment analysis and natural language processing in light of the large …


An Exploration Of Parliamentary Speeches In The Irish Parliament Using Topic Modeling, Fiona Leheny Jan 2018

An Exploration Of Parliamentary Speeches In The Irish Parliament Using Topic Modeling, Fiona Leheny

Dissertations

The only resource available in the public domain which highlights parliamentary ac tivity is parliamentary questions. Up until the last ten years, manual content analysis was carried out to classify these. More recently, machine learning techniques have been used to automatically classify and analyse these data sets. This study analyses the verbal parliamentary speeches in the Irish Parliament (known as the D´ail) over a ten year period using unsupervised machine learning. It does so by applying a less utilised topic modeling technique, known as Non-negative Matrix Factorisation (NMF), to de tect the latent themes in these speeches. A two-layer dynamic …


Automatic Log Parser To Support Forensic Analysis, Hudan Studiawan, Ferdous Sohel, Christian Payne Jan 2018

Automatic Log Parser To Support Forensic Analysis, Hudan Studiawan, Ferdous Sohel, Christian Payne

Australian Digital Forensics Conference

Event log parsing is a process to split and label each field in a log entry. Existing approaches commonly use regular expressions or parsing rules to extract the fields. However, such techniques are time-consuming as a forensic investigator needs to define a new rule for each log file type. In this paper, we present a tool, namely nerlogparser, to parse the log entries automatically, where log parsing is modeled as a named entity recognition problem. We use a deep machine learning technique, specifically the bidirectional long short-term memory networks, as the underlying architecture for this purpose. Unlike existing tools, nerlogparser …


River Report. State Of The Lower St. Johns River Basin, Florida: Water Quality, Fisheries, Aquatic Life, Contaminants, 2018, Environmental Protection Board, City Of Jacksonville, University Of North Florida, Jacksonville University, Gerry Pinto, Gretchen Bielmyer-Fraser, Nisse Goldberg, Anthony Ouellette, An-Phong Le, Radha Pyati, Charles Closmann, Brian P. Zoellner Jan 2018

River Report. State Of The Lower St. Johns River Basin, Florida: Water Quality, Fisheries, Aquatic Life, Contaminants, 2018, Environmental Protection Board, City Of Jacksonville, University Of North Florida, Jacksonville University, Gerry Pinto, Gretchen Bielmyer-Fraser, Nisse Goldberg, Anthony Ouellette, An-Phong Le, Radha Pyati, Charles Closmann, Brian P. Zoellner

State of the River Report

No abstract provided.


Front Matter Jan 2018

Front Matter

The Synapse: Intercollegiate science magazine

No abstract provided.


Sentiment Of The Union: Analyzing Tone In Presidential State Of The Union Addresses, Chase Rydeen Jan 2018

Sentiment Of The Union: Analyzing Tone In Presidential State Of The Union Addresses, Chase Rydeen

Honors Theses

As the machine learning and data science craze sweeps the nation, the implications and implementations are vast. This paper takes a look at both of them through the lens of a topic of national importance, at the very least for the United States. This topic is the words used by past Presidents of the United States, which are being pulled from their State of the Union Addresses. The focus of this research is on Natural Language Processing (NLP) and it's applied processes. Natural Language Processing allows for effective analysis of text-based data. Using NLP, a sentiment analysis was conducted on …


A Wikipedia Powered State-Based Approach To Automatic Search Query Enhancement, Kyle Goslin, Markus Hofmann Jan 2018

A Wikipedia Powered State-Based Approach To Automatic Search Query Enhancement, Kyle Goslin, Markus Hofmann

Articles

This paper describes the development and testing of a novel Automatic Search Query Enhancement (ASQE) algorithm, the Wikipedia N Sub-state Algorithm (WNSSA), which utilises Wikipedia as the sole data source for prior knowledge. This algorithm is built upon the concept of iterative states and sub-states, harnessing the power of Wikipedia's data set and link information to identify and utilise reoccurring terms to aid term selection and weighting during enhancement. This algorithm is designed to prevent query drift by making callbacks to the user's original search intent by persisting the original query between internal states with additional selected enhancement terms. The …


Mental Models And Neutrosophic Cognitive Maps (Modelos Mentales Y Mapas Cognitivos Neutrosóficos), Maykel Leyva-Vazquez, Rebeca Escobar-Jara, Florentin Smarandache Jan 2018

Mental Models And Neutrosophic Cognitive Maps (Modelos Mentales Y Mapas Cognitivos Neutrosóficos), Maykel Leyva-Vazquez, Rebeca Escobar-Jara, Florentin Smarandache

Branch Mathematics and Statistics Faculty and Staff Publications

In this work, elements related to mental models elicitation and analysis are addressed through causal models. Issues related to the need to include indeterminacy in causal relationships through neutrophic cognitive maps are discussed. A proposal for static analysis in neutrosophic cognitive maps is presented. The following activities are included in the proposal: Calculate, measures of centrality, Classify nodes, De-neutrosification, and Ranking nodes. As future works, the incorporation of new metrics of centrality in neutrosophic cognitive maps is proposed. The inclusion of scenario analysis to the proposal is another area of future work.


Understanding The Novice Decision-Making Process In Forensic Footwear Examinations: Accuracy And Decision Rules, Madonna A. Nobel Jan 2018

Understanding The Novice Decision-Making Process In Forensic Footwear Examinations: Accuracy And Decision Rules, Madonna A. Nobel

Graduate Theses, Dissertations, and Problem Reports (ETD)

The reproducibility of experienced-based forensic pattern interpretation is founded on the notion that domain-specific knowledge can be successfully distributed and applied among experts within a group. This assumption persists, even when the examination is complicated by variations in case circumstances, such as impression clarity and totality, as well as media, substrate, collection mechanism and enhancement. While it is further theorized that many of these factors (as well as additional confounding factors) are at play during an examination, the manner and extent to which these sources of variability affect the examination of footwear evidence remain unclear. In order to explore this …


Student Integration Into Stem Careers And Culture: A Longitudinal Examination Of Summer Faculty Mentors And Project Ownership, Michelle Richards-Babb Jan 2018

Student Integration Into Stem Careers And Culture: A Longitudinal Examination Of Summer Faculty Mentors And Project Ownership, Michelle Richards-Babb

Faculty & Staff Scholarship

It is widely recognized that the United States needs to attract and retain more people in science, technology, engineering, and mathematics (STEM) careers. Intensive undergraduate research experiences (UREs) are one of the few strategies shown to improve longitudinal student interest and persistence in STEM-related career pathways; however, less is known about the underlying process linking activities to positive outcomes. The tripartite integration model of social influences (TIMSI) provides a framework for understanding the social influence processes by which students integrate into STEM careers and culture. The current study used a longitudinal design and latent growth curve modeling to examine and …


Elephant 2000: A Programming Language For Remembering The Past And Building On It, Kerry J. Holmes Jan 2018

Elephant 2000: A Programming Language For Remembering The Past And Building On It, Kerry J. Holmes

Williams Honors College, Honors Research Projects

Elephant 2000 is a programming language to specify programs that accept user speech as text inputs and outputs speech text. The inputs and outputs are based on Dialogue Act theory which describes several forms of speech outputs, such as requests, questions, and answers. The language also relies on Named Entity Recognition to determine what types of objects a user references. These entities include persons, locations, times and so on. Using these attributes of user speech, a program is able to perform simple rule matching and pattern recognition to respond to input. The result is a programming language with English like …


Representation Learning With Convolutional Neural Networks, Haotian Xu Jan 2018

Representation Learning With Convolutional Neural Networks, Haotian Xu

Wayne State University Dissertations

Deep learning methods have achieved great success in the areas of Computer Vision and Natural Language Processing. Recently, the rapidly developing field of deep learning is concerned with questions surrounding how we can learn meaningful and effective representations of data. This is because the performance of machine learning approaches is heavily dependent on the choice and quality of data representation, and different kinds of representation entangle and hide the different explanatory factors of variation behind the data.

In this dissertation, we focus on representation learning with deep neural networks for different data formats including text, 3D polygon shapes, and brain …


Anatomy Of Online Hate: Developing A Taxonomy And Machine Learning Models For Identifying And Classifying Hate In Online News Media, Joni Salminen, Hind Almerekhi, Milica Milenkovic, Soon-Gyu Jung, Haewoon Kwak, Haewoon Kwak, Bernard J. Jansen Jan 2018

Anatomy Of Online Hate: Developing A Taxonomy And Machine Learning Models For Identifying And Classifying Hate In Online News Media, Joni Salminen, Hind Almerekhi, Milica Milenkovic, Soon-Gyu Jung, Haewoon Kwak, Haewoon Kwak, Bernard J. Jansen

Research Collection School Of Computing and Information Systems

Online social media platforms generally attempt to mitigate hateful expressions, as these comments can be detrimental to the health of the community. However, automatically identifying hateful comments can be challenging. We manually label 5,143 hateful expressions posted to YouTube and Facebook videos among a dataset of 137,098 comments from an online news media. We then create a granular taxonomy of different types and targets of online hate and train machine learning models to automatically detect and classify the hateful comments in the full dataset. Our contribution is twofold: 1) creating a granular taxonomy for hateful online comments that includes both …


Artificial Neural Network (Ann) In A Small Dataset To Determine Neutrality In The Pronunciation Of English As A Foreign Language In Filipino Call Center Agents, Proceso L. Fernandez Jr, Rey Benjamin M. Baquirin Jan 2018

Artificial Neural Network (Ann) In A Small Dataset To Determine Neutrality In The Pronunciation Of English As A Foreign Language In Filipino Call Center Agents, Proceso L. Fernandez Jr, Rey Benjamin M. Baquirin

Department of Information Systems & Computer Science Faculty Publications

Artificial Neural Networks (ANNs) have continued to be efficient models in solving classification problems. In this paper, we explore the use of an ANN with a small dataset to accurately classify whether Filipino call center agents’ pronunciations are neutral or not based on their employer’s standards. Isolated utterances of the ten most commonly used words in the call center were recorded from eleven agents creating a dataset of 110 utterances. Two learning specialists were consulted to establish ground truths and Cohen’s Kappa was computed as 0.82, validating the reliability of the dataset. The first thirteen Mel-Frequency Cepstral Coefficients (MFCCs) were …


A Twitter-Based Study For Understanding Public Reaction On Zika Virus, Roopteja Muppalla Jan 2018

A Twitter-Based Study For Understanding Public Reaction On Zika Virus, Roopteja Muppalla

Browse all Theses and Dissertations

In recent times, social media platforms like Twitter have become more popular and people have become more interactive and responsive than before. People often react to every news in real-time and within no-time, the information spreads rapidly. Even with viral diseases like Zika, people tend to share their opinions and concerns on social media. This can be leveraged by the health officials to track the disease in real-time thereby reducing the time lag due to traditional surveys. A faster and accurate detection of the disease can allow health officials to understand people's opinion of the disease and take necessary precautions …


The Feasibility Of Dementia Caregiver Task Performance Measurement Using Smart Gaming Technology, Garrett G. Goodman Jan 2018

The Feasibility Of Dementia Caregiver Task Performance Measurement Using Smart Gaming Technology, Garrett G. Goodman

Browse all Theses and Dissertations

Dementia caregiver burnout is detrimental to both the familial caregiver and their loved ones with dementia. As the population of older adults increases, both the number of individuals with dementia and their corresponding caregivers increase as well. Thus, we are interested in developing a potential tool to non-invasively detect signs of caregiver burnout using a mobile application combined with machine learning. Hence, the mobile application "Caregiver Assessment using Smart Technology" (CAST) was developed which personalizes a word scramble game. The CAST application utilizes a heuristically constructed Fuzzy Inference System (FIS) optimized via a Genetic Algorithm (GA) to provide an individualized …


Development Of A Human-Ai Teaming Based Mobile Language Learning Solution For Dual Language Learners In Early And Special Educations, Saurabh Shukla Jan 2018

Development Of A Human-Ai Teaming Based Mobile Language Learning Solution For Dual Language Learners In Early And Special Educations, Saurabh Shukla

Browse all Theses and Dissertations

Learning English as a secondary language is often an overwhelming challenge for dual language learners (DLLs), whose first language (L1) is not English, especially for children in early education (PreK-3 age group). These early DLLs need to devote a considerable amount of time learning to speak and read the language, in order to gain the language proficiency to function and compete in the classroom. Fear of embarrassment when mispronouncing words in front of others may drive them to remain silent; effectively hampering their participation in the class and overall curricular growth. The process of learning a new language can benefit …


A Framework To Understand Emoji Meaning: Similarity And Sense Disambiguation Of Emoji Using Emojinet, Sanjaya Wijeratne Jan 2018

A Framework To Understand Emoji Meaning: Similarity And Sense Disambiguation Of Emoji Using Emojinet, Sanjaya Wijeratne

Browse all Theses and Dissertations

Pictographs, commonly referred to as `emoji’, have become a popular way to enhance electronic communications. They are an important component of the language used in social media. With their introduction in the late 1990’s, emoji have been widely used to enhance the sentiment, emotion, and sarcasm expressed in social media messages. They are equally popular across many social media sites including Facebook, Instagram, and Twitter. In 2015, Instagram reported that nearly half of the photo comments posted on Instagram contain emoji, and in the same year, Twitter reported that the `face with tears of joy’ emoji has been tweeted 6.6 …


Human-Intelligence And Machine-Intelligence Decision Governance Formal Ontology, Faisal Mahmud Jan 2018

Human-Intelligence And Machine-Intelligence Decision Governance Formal Ontology, Faisal Mahmud

Engineering Management & Systems Engineering Theses & Dissertations

Since the beginning of the human race, decision making and rational thinking played a pivotal role for mankind to either exist and succeed or fail and become extinct. Self-awareness, cognitive thinking, creativity, and emotional magnitude allowed us to advance civilization and to take further steps toward achieving previously unreachable goals. From the invention of wheels to rockets and telegraph to satellite, all technological ventures went through many upgrades and updates. Recently, increasing computer CPU power and memory capacity contributed to smarter and faster computing appliances that, in turn, have accelerated the integration into and use of artificial intelligence (AI) in …


Improved Method Of Heuristic Classification Of Vowels From An Acoustic Signal, Josef Krocil, Zdenek Machacek, Jiri Koziorek, Radek Martinek, Jan Nedoma, Marcel Fajkus Jan 2018

Improved Method Of Heuristic Classification Of Vowels From An Acoustic Signal, Josef Krocil, Zdenek Machacek, Jiri Koziorek, Radek Martinek, Jan Nedoma, Marcel Fajkus

Turkish Journal of Electrical Engineering and Computer Sciences

This paper describes research in the field of the improved methodology of the classification of vowels /a, a:/, /$\varepsilon$, $\varepsilon$:/, /ı, i:/, /o, o:/, and /u, u:/ (vowel symbols according to IPA, i.e. International Phonetic Alphabet). The aim is to develop an improved method enabling the automatic allocation of vowel symbols to the corresponding time segments of acoustic recordings of an undisturbed speech signal. The combined classification method is based on finding frequencies of the first two local maxims (formants) in a smoothed linear predictive amplitude spectrum (LPC, linear predictive coding) and zero-crossing values of each speech active voiced short-term …


A Deep Relation Extraction From Biomedical Texts With Attention Mechanisms And Domain-Specific Contextual Representations, Amarin Jettakul Jan 2018

A Deep Relation Extraction From Biomedical Texts With Attention Mechanisms And Domain-Specific Contextual Representations, Amarin Jettakul

Chulalongkorn University Theses and Dissertations (Chula ETD)

The biomedical relation extraction (RE) tasks aim to study the interaction between pre-defined entities from biomedical literature: Bacteria Biotope (BB) and Drug-Drug interactions (DDI) tasks. Some previous investigations have used feature-based models; others have presented deep-learning-based models such as convolutional and recurrent neural networks used with the shortest dependency paths (SDPs). Although SDPs contain valuable and concise information, sections of significant information necessary to define bacterial location relationships are often neglected. In addition, the traditional word embedding used in previous studies may suffer from word ambiguation across linguistic contexts. Here, we present a deep learning model for biomedical RE. The …


การจำแนกเสียงโกรธในบทสนทนาของศูนย์ให้บริการข้อมูล, วิฎากร แซ่หว่อง Jan 2018

การจำแนกเสียงโกรธในบทสนทนาของศูนย์ให้บริการข้อมูล, วิฎากร แซ่หว่อง

Chulalongkorn University Theses and Dissertations (Chula ETD)

ศูนย์ให้บริการข้อมูลเป็นส่วนงานขององค์กรที่เกี่ยวข้องกับข้อมูลเสียงโดยตรง อารมณ์โกรธของลูกค้าที่เกิดขึ้นเป็นสิ่งที่ส่งผลในด้านลบต่อองค์กร งานในการตรวจสอบอารมณ์โกรธที่เกิดขึ้นให้ถูกต้องและรวดเร็วจึงเป็นสิ่งที่องค์กรให้ความสำคัญเนื่องจากมีจำนวนผู้ติดต่อเข้ามาใช้บริการจำนวนมากในแต่ละวัน จึงเป็นที่มาของการศึกษาการพัฒนาตัวแบบรู้จำอารมณ์โกรธของลูกค้าในบริบทของข้อมูลเสียงสนทนาระหว่างลูกค้าและเจ้าหน้าที่ของศูนย์ให้บริการข้อมูล การพัฒนาตัวแบบรู้จำอารมณ์มีขั้นตอนสำคัญ 4 ขั้นตอน ได้ แก่ (1) การประมวลผลข้อมูลเสียงเบื้องต้น เป็นการเตรียมข้อมูลเสียงให้พร้อมสำหรับการวิเคราะห์โดยแยกเสียงของลูกค้าออกจากสายสนทนา และตัดให้เป็นรอบการโต้ตอบ (2) การระบุอารมณ์ให้กับรอบการโต้ตอบ (3) การสกัดคุณลักษณะทางคลื่นเสียงจำนวน 1,582 คุณลักษณะที่สามารถใช้จำแนกอารมณ์ที่เกิดขึ้นภายในรอบการโต้ตอบ และ (4) การพัฒนาตัวแบบ โดยผู้ศึกษาได้เปรียบเทียบประสิทธิภาพของเทคนิค 4 เทคนิค ได้แก่ แรนดอมฟอร์เรส ดิซิสชั่นทรี ลิเนียร์เอสวีซีและโลจิสติกส์รีเกรสชั่น จากข้อมูลรอบการโต้ตอบทั้งหมดที่ใช้พัฒนาตัวแบบคือ 1,345 รอบการโต้ตอบ ที่แบ่งเป็นรอบการโต้ตอบที่เป็นอารมณ์โกรธจำนวน 517 รอบ และรอบการโต้ตอบที่ไม่พบอารมณ์โกรธหรือเป็นอารมณ์ปกติจำนวน 828 รอบ ผลการวิจัยพบว่า เทคนิคแรนดอมฟอร์เรสเป็นเทคนิคที่ให้ประสิทธิภาพที่ดีที่สุด โดยมีค่าระลึกของอารมณ์โกรธเท่ากับ 83.49 เปอร์เซ็นต์ และค่าความถูกต้องสมดุล 85.53 เปอร์เซ็นต์ นอกจากนั้นยังพบว่ารอบการโต้ตอบที่มีความยาวที่ 4 วินาทีขึ้นไปจะได้ผลลัพธ์ในการทำนายที่แม่นยำมากกว่าทั้งในรอบการโต้ตอบที่เป็นอารมณ์โกรธและอารมณ์ปกติ


การประยุกต์ใช้เทคนิคการเรียนรู้เชิงลึกในการฝังคำเพื่อสร้างแบบจำลองสำหรับทำนายผลการวินิจฉัยโรคจากบันทึกทางการแพทย์ของแผนกออร์โธปิดิกส์, ธนากร รัตนจริยา Jan 2018

การประยุกต์ใช้เทคนิคการเรียนรู้เชิงลึกในการฝังคำเพื่อสร้างแบบจำลองสำหรับทำนายผลการวินิจฉัยโรคจากบันทึกทางการแพทย์ของแผนกออร์โธปิดิกส์, ธนากร รัตนจริยา

Chulalongkorn University Theses and Dissertations (Chula ETD)

งานวิจัยนี้จะนำเสนอวิธีการฝังคำในการเรียนรู้เชิงลึกเพื่อสร้างแบบจำลองสำหรับทำนายผลการวินิจฉัยโรค ซึ่งปัจจัยสำคัญที่ทำให้ผลการวินิจฉัยโรคของแพทย์ผิดพลาดคือประสบการณ์ของแพทย์ที่ไม่เพียงพอ โดยการวินิจฉัยโรคที่ผิดพลาดนั้น นอกจากจะนำไปสู่การรักษาที่ผิดพลาดแล้ว ยังทำให้ผู้ป่วยเสียทั้งเงินและเวลา ดังนั้นเพื่อแก้ไขปัญหาการวินิจฉัยที่ผิดพลาด งานวิจัยนี้จึงนำเสนอวิธีการประยุกต์ใช้การเรียนรู้เชิงลึกกับการฝังคำ เพื่อทำนายผลการวินิจฉัยโรคจากระบบเวชระเบียน โดยจะสร้างแบบจำลองจากการใช้ข้อมูลในบันทึกของแพทย์ ซึ่งข้อมูลจะถูกนำไปวิเคราะห์ผ่านแบบจำลอง เพื่อทำนายผลการวินิจฉัยโรคที่มีความน่าจะเป็นออกมา เรียงตามลำดับความเชื่อมั่น และสุดท้ายจะใช้อัตราผลบวกจริง อัตราผลบวกเท็จ และค่าความแม่นยำมาเป็นตัววัดประสิทธิภาพของแบบจำลองที่ได้ ซึ่งพบว่าค่าความแม่นยำของแบบจำลองในงานวิจัยนี้มีค่าเท่ากับ 99.95% และอัตราผลบวกจริงมีค่าเท่ากับ 86.64% ด้วยการทำนายผลลัพธ์อันดับแรกเพียงอันดับเดียว


ระบบสังเคราะห์เสียงร้องเพลงภาษาไทยโดยใช้แบบจำลองฮิดเดนมาร์คอฟ, ลัทธพล จีระประดิษฐ Jan 2018

ระบบสังเคราะห์เสียงร้องเพลงภาษาไทยโดยใช้แบบจำลองฮิดเดนมาร์คอฟ, ลัทธพล จีระประดิษฐ

Chulalongkorn University Theses and Dissertations (Chula ETD)

การร้องเพลงในแต่ละภาษานั้นมีเอกลักษณ์เฉพาะตัวบางอย่างซึ่งส่งผลให้การพัฒนาความเป็นธรรมชาติของเสียงร้องเพลงสังเคราะห์ในแต่ละภาษานั้นมีความท้าทายแตกต่างกัน เสียงวรรณยุกต์เป็นส่วนที่มีผลมากกับการสื่อสารในภาษาที่มีเสียงวรรณยุกต์ แต่ระบบสังเคราะห์เสียงร้องเพลงในปัจจุบันไม่ได้คำนึงถึงเสียงวรรณยุกต์ นอกจากนี้เมลิสมาเป็นอีกสถานการณ์หนึ่งที่พบได้บ่อยครั้งในการร้องเพลงป็อปไทยซึ่งต้องมีการจัดการเพื่อจำลองการร้องเมลิสมา เป้าหมายของวิทยานิพนธ์นี้จึงมุ่งเน้นที่การปรับระบบสังเคราะห์เสียงร้องเพลงให้รองรับการจำลองเสียงในสถานการณ์เมลิสมาและผลกระทบของเสียงวรรณยุกต์ งานวิทยานิพนธ์นี้เสนอ 1) ปัจจัยบริบทที่ใช้ในระบบสังเคราะห์เสียงร้องเพลงสำหรับภาษาที่วรรณยุกต์มีผลต่อเสียงร้องเพลงและคำนึงถึงเมลิสมา 2) วิธีการทำสำเนารูปเขียน จากการประเมินผลพบว่า วิธีการทำสำเนารูปเขียนที่เสนอทั้งสองแบบนั้นส่งผลให้ระบบสังเคราะห์เสียงร้องเพลงรองรับเมลิสมา โดยวิธีการทำสำเนารูปเขียนที่คำนึงถึงสระเสียงสั้น-ยาวและตัวสะกดนั้นมีรูปคลื่นของเสียงร้องเพลงสังเคราะห์ใกล้เคียงกับรูปคลื่นของเสียงร้องเพลงจริงมากกว่า รวมถึงมีความเป็นธรรมชาติมากกว่าโดยใช้มาตรวัดเอ็มโอเอส อีกทั้งเมื่อมีปัจจัยบริบทที่เกี่ยวข้องกับเสียงวรรณยุกต์ เค้ารูปของความถี่มูลฐานที่สังเคราะห์ได้นั้นมีความใกล้เคียงเสียงร้องเพลงจริงมากกว่าในระบบที่ไม่มีปัจจัยบริบทที่เกี่ยวข้องกับเสียงวรรณยุกต์ และมีความเป็นธรรมชาติมากขึ้นโดยใช้มาตรวัดเอ็มโอเอส นอกจากนี้เพื่อเพิ่มความเป็นธรรมชาติให้เสียงร้องเพลงสังเคราะห์จึงมีการทดลองเกี่ยวกับจำนวนสถานะของแบบจำลองเสียงพบว่า เมื่อจำนวนสถานะเพิ่มขึ้น ความเป็นธรรมชาติของเสียงร้องเพลงสังเคราะห์ก็มากขึ้น แต่เมื่อถึงจุดหนึ่งเสียงร้องเพลงสังเคราะห์ที่ได้จะมีความเป็นธรรมชาติลดลง


Using Semantic And Pragmatic Information To Determine Syntactic Decisions In Translation Of The Passive Voice From English To French, Ashli Fain Jan 2018

Using Semantic And Pragmatic Information To Determine Syntactic Decisions In Translation Of The Passive Voice From English To French, Ashli Fain

Graduate Research Theses & Dissertations

I investigated statistically significant factors that effect syntactic changes in French and

English during translation of by phrases using the passive voice. Relative clauses containing this

structure were also looked at. This feature was chosen from information that was gathered from

experts in the form of academic texts on English-French translation, as well as handbooks

intended for an audience of professional translators. Data came from the proceedings of the

Canadian Parliament (the Hansard), which are published in French and English. Passive

sentences were collected from the output of the Stanford Parser. Animacy was found to be a

significant factor in …


Tracking Topical Evolution In Large Document Collections, Sheikh Motahar Naim Jan 2018

Tracking Topical Evolution In Large Document Collections, Sheikh Motahar Naim

Open Access Theses & Dissertations

A large document collection that builds up over time usually contains a number of different themes. All of these themes or topics are not equally important at the same time. One topic might have high probabilities in some years due to some relevant events, and low probabilities in other years. Analyzing the evolution of such topics has useful applications in a variety of domains, for example, helping researchers to quickly see the changes of research topics in an area, assisting intelligence agents in tracking the activities of a terrorist group, or monitoring damages caused by a natural disaster. In this …