Semantic Feature Extraction For Narrative Analysis

A story is defined as "an actor(s) taking action(s) that culminates in a resolution(s)''. I present novel sets of features to facilitate story detection among text via supervised classification and further reveal different forms within stories via unsupervised clustering. First, I investigate the utility of a new set of semantic features compared to standard keyword features combined with statistical features, such as density of part-of-speech (POS) tags and named entities, to develop a story classifier. The proposed semantic features are based on Subject, Verb, Object triplets that can be extracted using a shallow parser. Experimental results show that a model of memory-based semantic linguistic features alongside statistical features achieves better accuracy. Next, I further improve the performance of story detection with a novel algorithm which aggregates the triplets producing generalized concepts and relations. A major challenge in automated text analysis is that different words are used for related concepts. Analyzing text at the surface level would treat related concepts (i.e. actors, actions, targets, and victims) as different objects, potentially missing common narrative patterns. The algorithm clusters Subject, Verb, Object triplets into generalized concepts by utilizing syntactic criteria based on common contexts and semantic corpus-based statistical criteria based on "contextual synonyms''. Generalized concepts representation of text (1) overcomes surface level differences (which arise when different keywords are used for related concepts) without drift, (2) leads to a higher-level semantic network representation of related stories, and (3) when used as features, they yield a significant (36%) boost in performance for the story detection task. Finally, I implement co-clustering based on generalized concepts/relations to automatically detect story forms. Overlapping generalized concepts and relationships correspond to archetypes/targets and actions that characterize story forms. I perform co-clustering of stories using standard unigrams/bigrams and generalized concepts. I show that the residual error of factorization with concept-based features is significantly lower than the error with standard keyword-based features. I also present qualitative evaluations by a subject matter expert, which suggest that concept-based features yield more coherent, distinctive and interesting story forms compared to those produced by using standard keyword-based features.

The objective of this monograph is to improve the performance of the sentiment analysis model by incorporating the semantic, syntactic and common-sense knowledge. This book proposes a novel semantic concept extraction approach that uses dependency relations between words to extract the features from the text. Proposed approach combines the semantic and common-sense knowledge for the better understanding of the text. In addition, the book aims to extract prominent features from the unstructured text by eliminating the noisy, irrelevant and redundant features. Readers will also discover a proposed method for efficient dimensionality reduction to alleviate the data sparseness problem being faced by machine learning model. Authors pay attention to the four main findings of the book : -Performance of the sentiment analysis can be improved by reducing the redundancy among the features. Experimental results show that minimum Redundancy Maximum Relevance (mRMR) feature selection technique improves the performance of the sentiment analysis by eliminating the redundant features. - Boolean Multinomial Naive Bayes (BMNB) machine learning algorithm with mRMR feature selection technique performs better than Support Vector Machine (SVM) classifier for sentiment analysis. - The problem of data sparseness is alleviated by semantic clustering of features, which in turn improves the performance of the sentiment analysis. - Semantic relations among the words in the text have useful cues for sentiment analysis. Common-sense knowledge in form of ConceptNet ontology acquires knowledge, which provides a better understanding of the text that improves the performance of the sentiment analysis.

This book constitutes the refereed proceedings of the 9th Metadata and Semantics Research Conference, MTSR 2015, held in Manchester, UK, in September 2015.The 35 full papers and 3 short papers presented together with 2 poster papers were carefully reviewed and selected from 76 submissions. The papers are organized in several sessions and tracks: general track on ontology evolution, engineering, and frameworks, semantic Web and metadata extraction, modelling, interoperability and exploratory search, data analysis, reuse and visualization; track on digital libraries, information retrieval, linked and social data; track on metadata and semantics for open repositories, research information systems and data infrastructure; track on metadata and semantics for agriculture, food and environment; track on metadata and semantics for cultural collections and applications; track on European and national projects.

An important recent development in the study of teaching is the use of narrative analysis to study teachers' lives, their work and anecdotes exchanged in the staffroom.; This book critically examines current approaches to the study of teachers' narratives and argues that, for narrative research to be effective, we need to see narrative in a multi- disciplinary perspective. The book examines models of narrative analysis currently proposed in linguistics, sociology, psycology, anthropology and literature and applies insights from these disciplines to the study of teachers' narratives. The author proposes an alternative approach to studying narratives which is then applied to original data, demonstrating how narrative analysis can be used to study primary teachers' perceptions of their work. lt is suggested that narrative analysis could be used to study the perceptions or culture of any professional group.

This book constitutes the thoroughly refereed proceedings of the 11th International Conference on Metadata and Semantic Research, MTSR 2017 2017, held in Tallinn, Estonia, November 28th to December 1st, 2017. The 18 full and 13 short papers presented were carefully reviewed and selected from 58 submissions. They focus on the Internet of Things (IoT) and the practical implementation of ontologies and linked data. Further topics are theoretical and foundational principles of metadata; ontologies and information organization; applications of linked data, open data, big data and user-generated metadata; digital interconnectedness; metadata standardization; authority control and interoperability in digital libraries and research data repositories; emerging issues in RDF, OWL, SKOS, schema.org, BIBFRAME, metadata and ontology design; linked data applications for e-books; digital publishing and Content Management Systems (CMSs); content discovery services, search, information retrieval and data visualization applications.

International Conference on Bio-Inspired Computing: Theories and Applications (BIC-TA) is one of the flagship conferences on Bio-Computing, bringing together the world’s leading scientists from different areas of Natural Computing. Since 2006, the conferences have taken place at Wuhan (2006), Zhengzhou (2007), Adelaide (2008), Beijing (2009), Liverpool & Changsha (2010), Malaysia (2011) and India (2012). Following the successes of previous events, the 8th conference is organized and hosted by Anhui University of Science and Technology in China. This conference aims to provide a high-level international forum that researchers with different backgrounds and who are working in the related areas can use to present their latest results and exchange ideas. Additionally, the growing trend in Emergent Systems has resulted in the inclusion of two other closely related fields in the BIC-TA 2013 event, namely Complex Systems and Computational Neuroscience. These proceedings are intended for researchers in the fields of Membrane Computing, Evolutionary Computing and Genetic Algorithms, DNA and Molecular Computing, Biological Computing, Swarm Intelligence, Autonomy-Oriented Computing, Cellular and Molecular Automata, Complex Systems, etc. Professor Zhixiang Yin is the Dean of the School of Science, Anhui University of Science & Technology, China. Professor Linqiang Pan is the head of the research group of Natural Computing at Huazhong University of Science and Technology, Wuhan, China. Professor Xianwen Fang also works at the Anhui University of Science & Technology.

Online communities generate massive volumes of natural language data and the social sciences continue to learn how to best make use of this new information and the technology available for analyzing it. Text Mining: A Guidebook for the Social Sciences brings together a broad range of contemporary qualitative and quantitative methods to provide strategic and practical guidance on analyzing large text collections. This accessible book, written by sociologist Gabe Ignatow and computer scientist Rada Mihalcea, surveys the fast-changing landscape of data sources, programming languages, software packages, and methods of analysis available today. Suitable for novice and experienced researchers alike, the book will help readers use text mining techniques more efficiently and productively.

This extensively revised new edition comprehensively reviews the rise of clinical research informatics (CRI). It enables the reader to develop a thorough understanding of how CRI has developed and the evolving challenges facing the biomedical informatician in the modern clinical research environment. Emphasis is placed on the changing role of the consumer, and the need to merge clinical care delivery and research as part of a changing paradigm in global healthcare delivery. Clinical Research Informatics presents a detailed review of using informatics in the continually evolving clinical research environment. It represents a valuable textbook reference for all students and practising healthcare informaticians looking to learn and expand their understanding of this fast-moving and increasingly important discipline.

This study investigates the richly diverse but integrated semiotic potential of storytelling. Unlike other interdisciplinary approaches to narrative studies which have privileged the study of words in storytelling, this unique collection provides a much needed analysis of how narrative operates using combinations of visual, typographic, aural, gestural and haptic resources. Although both multimodal theory and narrative studies have been invigorated by a variety of theoretical approaches, this volume seeks to avoid a single dominant paradigm. Instead, the contributors use literary criticism, linguistics and new media frameworks in a series of critical studies that are directly engaged with a range of multimodal stories. The contributors analyze works that include oral accounts of personal experience, opera, cartoons, print literature and new media forms of storytelling such as experimental digital fiction and fanfiction.