AIML Knowledge Base Construction from Text Corpora

Same benefits, different communication patterns: Comparing Children's reading with a conversational agent vs. a human partner

Purpose Human–agent interaction (HAI) is increasingly influencing our personal and work lives through the proliferation of conversational agents (CAs) in various domains. As such, these agents combine intuitive natural language interactions by also delivering personalization through artificial intelligence capabilities. However, research on CAs as well as practical failures indicates that CA interaction oftentimes fails miserably. To reduce these failures, this paper introduces the concept of building common ground for more successful HAIs. Design/methodology/approach Based on a systematic literature analysis, we identified 38 articles meeting the eligibility criteria. We critically reviewed this body of knowledge within a formal narrative synthesis structured around the use of common ground in the interaction with CAs. Findings Based on the systematic review, our analysis reveals five mechanisms for achieving common ground: embodiment, social features, joint action, knowledge base and mental model of conversational agent. We point out the relationships between these mechanisms as they are related to each other in directional and bidirectional ways. Research limitations/implications Our findings contribute to theory with several implications for CA research. First, we provide implications about the organization of common ground mechanisms for CAs. Second, we provide insights into the mechanisms and nomological network for achieving common ground when interacting with CAs. Third, we provide a broad research agenda for future CA research that centers around the important topic of common ground for HAI. Originality/value We offer novel insights into grounding mechanisms and highlight the potentials when considering common ground in different HAI processes. Consequently, we secure further understanding and deeper insights of possible mechanisms of common ground to shape future HAI processes.

The annual workshop on multiword expressions takes place since 2001 in conjunction with major computational linguistics conferences and attracts the attention of an ever-growing community working on a variety of languages, linguistic phenomena and related computational processing issues. MWE 2017 took place in Valencia, Spain, and represented a vibrant panorama of the current research landscape on the computational treatment of multiword expressions, featuring many high-quality submissions. Furthermore, MWE 2017 included the first shared task on multilingual identification of verbal multiword expressions. The shared task, with extended communal work, has developed important multilingual resources and mobilised several research groups in computational linguistics worldwide. This book contains extended versions of selected papers from the workshop. Authors worked hard to include detailed explanations, broader and deeper analyses, and new exciting results, which were thoroughly reviewed by an internationally renowned committee. We hope that this distinctly joint effort will provide a meaningful and useful snapshot of the multilingual state of the art in multiword expressions modelling and processing, and will be a point point of reference for future work.

The status of certain problems of statistical inference that have been encountered in mathematical and computational linguistics is described. Underlying statistical theory and techniques are illustrated by a variety of examples from structural, comparative, and historical linguistics. Concurrently, the relationships of some theoretical issues in statistical linguistics to practical matters in computational linguistics are briefly treated. Also, several areas in mathematical and computational linguistics of promising research interest for computer or information scientists are noted. Finally, the need for more interdisciplinary studies involving computer scientists, linguists, and mathematicians is pointed out.

Multi-word expressions are linguistic objects formed by two or more words that behave like a ‘unit’ by displaying formal and/or functional idiosyncratic properties with respect to free word combinations. They include an extremely varied set of items (from idioms to collocations, from formulae to sayings) which have been the privileged subject matter of fields such as phraseology, lexicology, lexicography, and computational linguistics. Far from being a marginal phenomenon, multi-word expressions are ubiquitous and pervasive: some estimate that they are as numerous as words in some languages, which makes them as central an issue as words for the understanding of human language. However, their relation with words, and morphology, is by far less explored, not to say neglected, especially in terms of demarcation, competition, and cross-linguistic variation.

Over the past two decades or so, Multi-Word Expressions (MWEs; also called Multi-word Units) have been an increasingly important concern for Computational Linguistics and Natural Language Processing (NLP). The term MWE has been used to refer to various types of linguistic units and expressions, including idioms, noun compounds, phrasal verbs, light verbs and other habitual collocations. However, while there is no universally agreed definition for MWE as yet, most researchers use the term to refer to those frequently occurring phrasal units which are subject to certain level of semantic opaqueness, or non-compositionality. Non-compositional MWEs pose tough challenges for automatic analysis because their interpretation cannot be achieved by directly combining the semantics of their constituents, thereby causing the "pain in the neck of NLP" (Sag et al. 2001). In fact, MWEs have been studied for decades in Phraseology under the term phraseological unit. But in the early 1990s, MWEs started receiving increasing attention in corpus-based computational linguistics and NLP. Early influential work on MWEs includes Smadja (1993), Dagan and Church (1994), Wu (1997), Daille (1995), Wermter and Chen (1997), McEnery et al. (1997), and Michiels and Dufour (1998). These studies address the automatic treatment of MWEs and their applications in practical NLP and information systems. A milestone for MWE

Conversational agents have been established to support educational practices, solving students’ questions, recommending teaching materials, increasing the motivation of students who have little interest in certain content, among others. Although many studies are discussing the establishing of conversational agents, there is no global understanding of what has been investigated in this area. There are some previous secondary studies that have attempted to map the conversational agents used for educational purposes (i.e., pedagogical conversational agents). However, these studies are limited to include primary studies published in specific sets of conferences and journals. Therefore, an overview of the current state of the art associated with pedagogical conversational agents has not been produced yet. This research paper aims to contribute to the theme, through the presentation of planning, conduction, and results analysis of a systematic mapping on pedagogical conversational agents. We address the following questions: (i) in which areas have conversation agents been investigated?; (ii) at which education levels are conversational agents used?; (iii) which mechanisms do conversational agents use to interact with students?. We also present some discussions about research opportunities that can be explored in future works. This study aims to contribute to the topic of pedagogical conversational agents, helping researchers to have a deeper understanding of what has already been done and to guide their research to subjects not explored yet.

Nowadays, the big challenge for community of Information Retrieval domain is to discover the relevance feature in text document which helps to decide whether document is relevant or irrelevant. Most existing text mining methods are based on term-based approaches which extract terms from a training set for describing relevant feature. In termbased approach multiple meaning of same word in different context leads to generate polysemy and synonymy issue. The term based approach also suffers from low level support problem. Even though pattern based text mining approach solve low level support problem but still this approach suffers from large number of noise pattern. In the propose work, a pattern discovery approach for text mining is explored. This approach discovers frequent sequential pattern and closed sequential patterns in text documents for identifying the most information contents of the documents and extract useful features for text mining. It also classifies extracted terms into three categories: positive terms, general terms, and negative terms. In this way, we update extracted features using multiple revising strategies. This technique discovers positive and negative patterns in text documents as higher level features in order to accurately weight low-level terms based on their specificity.

PurposeIn business, online reviews have an economic impact on firm performance. Customers’ data in the form of online reviews was used to understand the appreciation and service complaints written by previous customers. The study is an analysis of the online reviews written by the customers about Indian heritage hotels. This study aims to understand the dimensions of service appreciation and service complaints by comparing positive- and negative-rated reviews and find the patterns in the determinants of the satisfaction and dissatisfaction of the customers.Design/methodology/approachA total of 23,643 online reviews about heritage hotels were collected from the TripAdvisor website by using a Web crawler developed in Python. A total of 1000 reviews were randomly selected for further analysis to eliminate the bandwagon effect. Unsupervised text mining techniques were used to analyze reviews and find out the interesting patterns in text data.FindingsBased on Herzberg two-factor theory, this study found satisfied and dissatisfied determinants separately. The study revealed some common categories discussed by satisfied and dissatisfied customers. The factors which satisfy the customers may also dissatisfy the customers if not delivered properly. Satisfied customers mentioned about tangible features of the hotel stay, which includes physical signifiers, traditional services, staff behavior and professionalism and core products (rooms, food). However, most of the customers complained about intangible service problems, such as staff attitude, services failure, issues with reservation and food, value for money and room condition. The results are contradicting with commercial hotels-based studies owing to the unique services provided by heritage hotels.Practical implicationsThe dimensions for satisfaction and dissatisfaction among customer of heritage hotels provide marketers to understand the real emotion and perception of the customers. As these dimensions were extracted through text mining of the reviews written by the customer of heritage hotels, the results would certainly give better insights to the hotel marketers.Originality/valueThe study is a rare attempt to study online reviews of customers on heritage hotels through a text mining approach and find the patterns in the behavior and the determinants of satisfaction and dissatisfaction of customers.

News is new information or information about something that is currently happening, presented in the form of print media, broadcasts, social media or word of mouth to third parties or multiple people. The spread of this news is now very much happening both from social media or messages. However, the problem is that no one can guarantee the truth of the news received. Text mining is the application of data mining concepts and techniques to look for patterns in text, the process of analyzing text to find useful information for a particular purpose. Another definition related to text mining is that text mining is data mining in the form of text where the data source is usually obtained from documents and the goal is to find words that can represent the contents of the document so that an analysis of the connectivity between documents can be carried out. To make it easier to identify news hoax requires a text mining algorithm, using a text mining algorithm can be useful to get real news. Based on the calculations of the TF-IDF algorithm, it can be concluded that the similarity of the meaning of the news whose truth is not yet known is closest to the news ofmerahputih.com with a confidence level of 23.11%.

In Slovene linguistics, much research in phraseology has either been theoretical in nature or focused more on compiling lexicographic resources for human users. While several machine-readable lexicographic resources containing multiword expressions (MWEs) have also been developed in recent years, Slovene phraseology and computational Slovene linguistics remain largely divided into separate tracks. We attempt to bridge the gap with a brief demonstration of the benefits that computational and statistical approaches based on machine-readable data can have for linguists and phraseologists. We briefly present the SUK Training Corpus of Slovene, the largest machine-readable dataset for Slovene that contains annotations of multiword expressions, as well as the Q-CAT Corpus Annotation Tool that was used to annotate it. We extract examples for two Slovene MWEs (priti na zeleno vejo and podirati se kot hišica iz kart) from the morphosyntactically annotated Gigafida 2.1 Corpus of Written Standard Slovene using a rule-based approach that leverages syntactic structures. We perform a statistical analysis to determine the degree of variation within the extracted examples. We aim to show that machine-readable data is intended not only for developers of NLP tools but can also help provide additional insight into the structure and variation for the linguistic description of MWEs.

This book provides an up-to-date snapshot of recent research and developments in the use of corpora for language learning and teaching. It is divided into three parts. Part I focusses on innovative uses of corpora by language teachers and learners. These cover the world’s first corpus-based TV program for the teaching of English conversation, as well as corpus-based approaches to the teaching of EAP, cultural studies and translation. Part II focuses on new corpus-based tools for LSP learning. Part III illustrates research findings from corpora consisting of language learner data and discusses their implications for language teaching and learning. It will appeal to scholars in both language teaching and learning and corpus and computational linguistics.

Gigantic growth of text mining is becoming a potential source of crowd wisdom extraction and analysis especially in terms of text pre-processing and analysis. The ubiquitous growth in the text mining field is unprecedented where social media mining is playing a significant role. Therefore, it’s a significant task of current research purview which not only provides an appropriate platform for the contributors but also for active information seekers. Text mining has gained the widespread attention of internet users which is almost on an explosion level. Accessibility of data on such a behemoth scale mandates regular and critical analysis of this information for various perspectives’ plausibility. Rich patterns of online social text can be exploited to extract relevant information effectively. As text mining plays a significant and crucial role in discovery of these insights, therefore its challenges and contribution to social media analysis must be explored extensively. In this chapter, we provide a brief summary of the modern state of text mining in pre-processing and machine learning mechanism for COVID-19. Emphasis is placed on the resources and learning mechanism available to researchers and practitioners, as well as the major text mining tasks of interest to the community. Thus, the main aim of this chapter is to expound and intellectualize the domains of social media which are accessible on an extraordinary range in the field of text mining for COVID-19.

BackgroundVast numbers of domestic violence (DV) incidents are attended by the New South Wales Police Force each year in New South Wales and recorded as both structured quantitative data and unstructured free text in the WebCOPS (Web-based interface for the Computerised Operational Policing System) database regarding the details of the incident, the victim, and person of interest (POI). Although the structured data are used for reporting purposes, the free text remains untapped for DV reporting and surveillance purposes.ObjectiveIn this paper, we explore whether text mining can automatically identify mental health disorders from this unstructured text.MethodsWe used a training set of 200 DV recorded events to design a knowledge-driven approach based on lexical patterns in text suggesting mental health disorders for POIs and victims.ResultsThe precision returned from an evaluation set of 100 DV events was 97.5% and 87.1% for mental health disorders related to POIs and victims, respectively. After applying our approach to a large-scale corpus of almost a half million DV events, we identified 77,995 events (15.83%) that mentioned mental health disorders, with 76.96% (60,032/77,995) of those linked to POIs versus 16.47% (12,852/77,995) for the victims and 6.55% (5111/77,995) for both. Depression was the most common mental health disorder mentioned in both victims (22.25%, 3269) and POIs (18.70%, 8944), followed by alcohol abuse for POIs (12.19%, 5829) and various anxiety disorders (eg, panic disorder, generalized anxiety disorder) for victims (11.66%, 1714).ConclusionsThe results suggest that text mining can automatically extract targeted information from police-recorded DV events to support further public health research into the nexus between mental health disorders and DV.

The quality of extracted features is the key issue to text mining due to the large number of terms, phrases, and noise. Most existing text mining methods are based on term-based approaches which extract terms from a training set for describing relevant information. However, the quality of the extracted terms in text documents may be not high because of lot of noise in text. For many years, some researchers make use of various phrases that have more semantics than single words to improve the relevance, but many experiments do not support the effective use of phrases since they have low frequency of occurrence, and include many redundant and noise phrases. In this paper, we propose a novel pattern discovery approach for text mining. This approach first discovers closed sequential patterns in text documents for identifying the most informative contents of the documents and then utilise the identified contents to extract useful features for text mining. We develop a novel fusion method based on Dempster-Shafer's evidential reasoning which allows to combine the pieces of document to discover the knowledge (features). To evaluate the proposed approach, we adopt the feature extraction method for information filtering (IF). The experimental results conducted on Reuters Corpus Volume 1 and TREC topics confirm that the proposed approach could achieve excellent performance.