Text Mining

Natural Language Processing is a programmed approach to analyze text that is based on both a set of theories and a set of technologies. This forum aims to bring together researchers who have designed and build software that will analyze, understand, and generate languages that humans use naturally to address computers.

In recent years, organizations have come across complex databases due to the development of technology, the growth of databases and information technologies, and the use of widespread information technologies. If the data found in line with the needs of the institutions are managed successfully and effectively, it is obvious that the institutions and organizations will offer great advantages and opportunities in economic terms. Each action performed in the digital environment leaves a data record behind it. For this purpose, big data, science, data mining and text mining concepts, methods and usage areas were mentioned, and similar and different aspects of data mining and text mining were determined. Through this study, it is expected that the separation of data mining and text mining methods will accelerate the process of strategic decisions of institutions and organizations and provide the required support.

Automatic understanding of domain specific texts in order to extract useful relationships for later use is a nontrivial task. One such relationship would be between railroad accidents' causes and their correspondent descriptions in reports. From 2001 to 2016 rail accidents in the U.S. cost more than $4.6B. Railroads involved in accidents are required to submit an accident report to the Federal Railroad Administration (FRA). These reports contain a variety of fixed field entries including primary cause of the accidents (a coded variable with 389 values) as well as a narrative field which is a short text description of the accident. Although these narratives provide more information than a fixed field entry, the terminologies used in these reports are not easy to understand by a non-expert reader. Therefore, providing an assisting method to fill in the primary cause from such domain specific texts (narratives) would help to label the accidents with more accuracy. Another important question for transportation safety is whether the reported accident cause is consistent with narrative description. To address these questions, we applied deep learning methods together with powerful word embeddings such as Word2Vec and GloVe to classify accident cause values for the primary cause field using the text in the narratives. The results show that such approaches can both accurately classify accident causes based on report narratives and find important inconsistencies in accident reporting 1 .

The increasing number of digitized texts presently available notably on the Web has developed an acute need in text mining techniques. Clustering systems are used more and more often in text mining, especially to analyze texts and to extract knowledge they contain. With the availability of the vast amount of clustering algorithms and techniques, it becomes highly confusing to a user to choose the algorithm that best suits its target dataset. Actually, it is very hard to define which algorithms work the best, since results depend considerably on the application and on the kinds of data at hand. In this paper, we propose, study and compare three text clustering methods: an ascending hierarchical clustering method, a SOM-based clustering method and an ant-based clustering method, all of these based on the synsets of WordNet as terms for the representation of textual documents. The effects of these methods are examined in several experiments using 3 similarity measurements: the cosine distance, the Euclidean distance and the manhattan distance. The reuters-21578 corpus is used for evaluation. The evaluation was done, by using the F-measure. The results obtained show that the SOM-based clustering method using the cosine distance provides the best results.

In this paper, we analyze about the relation between stock price returns and Headline News. Headline News is very important sources of information in asset management, and is sent in large quantities every day. We study the effect of more than 13,000 Headline News sent from JIJI PRESS. We classify Headline News using Text Categorization and analyze the reaction of a stock price return for every type of News. From our research, we figure out following issues; 1) we make the Text Categorization System that has about 80% of classification accuracy, 2) this system can extract effective information to stock price returns from Headline News.

Co-clustering has been defined as a way to organize simultaneously subsets of instances and subsets of features in order to improve the clustering of both of them. In previous work [1], we proposed an efficient co-similarity measure allowing to simultaneously compute two similarity matrices between objects and features, each built on the basis of the other. Here we propose a generalization of this approach by introducing a notion of pseudo-norm and a pruning algorithm. Our experiments show that this new algorithm significantly improves the accuracy of the results when using either supervised or unsupervised feature selection data and that it outperforms other algorithms on various corpora.

Blogosphere is the name associated to universe of all the blog sites. A blog is a website that allows people to write about topics they want to share with others. The ease & simplicity of creating blog posts and their free form and unedited nature have made blogging a happening thing. The blogosphere today contains a large number of posts on virtually every topic of interest. This rich and unique source of data has attracted people and companies across disciplines to exploit it for varied purposes. However, it not only provides a rich data source for commercial exploitation but also for psychological & socio-political analysis purposes. This paper tries to identify the methodology and technical framework required for an analytical experiment of this kind, and demonstrates the plausibility of this idea through our clustering & opinion mining experiment on analysis of blog posts about a recent socio-political issue.

Analysis and modeling of crime text report data has important applications, including refinement of crime classifications, clustering of documents, and feature extraction for spatiotemporal forecasts. Having better neural network representations
of crime text data may facilitate all of these tasks. This paper evaluates the ability of generative adversarial network models to represent crime text data and generate realistic crime reports. We compare four state-of-the-art GAN algorithms in terms of quantitative metrics such as coherence, embedding similarity, negative log-likelihood, and qualitatively based on inspection of generated text. We discuss current challenges with crime text
representation and directions for future research

The objective of this research is to compare movement of stock return literature on companies in the tourism sector and other businesses. This study also analyzed comprehensively the relationship between contexts built on the literature. The mix method approach is adopted by combining text mining methods, which are analyzed using network and content analysis methods. The scientific literature was collected using the web scraping method on Litmaps and selected 200 articles and only used 195 articles (between 2000-2021) related to the topic of stock returns and 23 (between 2013-2020) related to the tourism company sector. In the process of data analysis and visualization, this study uses AntConc, RStudio, and Gephi software. This review discovered that in terms of topic and context, stock return studies discuss issues related to the influence of fundamental factors on stock returns, and vice versa, while the approach used is mostly quantitative based on the financial performance reports. Regarding the frequency of appearance, this result revealed that the effects of fundamental factors on stock returns in tourism companies can be translated into a number of strategic issues, including Current Ratio, TATO, ROE, FATO, and Systematic Risk, which provide a significant and positive effect on Direct Share Prices. Furthermore, the Current Ratio, TATO, ROE, and FATO are thought to have a significant and positive effect on stock returns via Systematic Risk. The scope of this research is limited to the
national journal (Indonesia) on the Litmaps search engine. In theory, the results of this research can be used as a comparison and considerations related to previous studies, as well as what investigation requires to be done immediately, in the context of tourism companies' stock returns. The literature study on the tourism companies’ stock returns issue is perhaps the first systematic literature review in Indonesia.

The textual analysis has become most important task due to the rapid increase of the number of texts that have been continuously generated in several forms such as posts and chats in social media, emails, articles, and news. The management of these texts requires efficient and
effective methods, which can handle the linguistic issues that come from the complexity of natural languages. In recent years, the exploitation of semantic features from the lexical sources has been widely investigated by researchers to deal with the issues of “synonymy and ambiguity” in the tasks involved in the Social Media like document clustering. The main challenges of exploiting the lexical knowledge sources such as 1WordNet 3.1 in these tasks are how to integrate
the various types of semantic relations for capturing additional semantic evidence, and how to settle the high dimensionality of current semantic representing approaches. In this paper, the proposed weighting of features for a new semantic feature-based method as which combined
four things as which is “Synonymy, Hypernym, non-taxonomy, and Glosses”. Therefore, this research proposes a new knowledge-based semantic representation approach for text mining, which can handle the linguistic issues as well as the high dimensionality issue. Thus, the
proposed approach consists of two main components: a feature-based method for incorporating the relations in the lexical sources, and a topic-based reduction method to overcome the high dimensionality issue. The proposed method approach will evaluated using WordNet 3.1 in the text clustering and text classification.

Analysis, Collecting and analyzing online tourist reviews on destinations is important for sustainable tourism. These analyses can give insight into the extent to which natural and cultural assets in the destination are protected. These evaluations should be considered by the authorities as objective and realistic assessments. In this study, 4183 TripAdvisor reviews of foreign tourists visiting "Göreme National Park and Cappadocia Rocky Area", which is listed in the World Heritage Site, were evaluated. The data set consisted of English reviews of foreign tourists visiting the region between the years of 2018-2020. Dictionary-based sentiment analysis, one of the text mining methods, was used in the study. According to the analysis results, the positive perceptions of the tourists about the churches, fairy chimneys, valleys and underground cities in the World Heritage Site were found to be significantly high (75%). Negative evaluations were found to be low (33%). In tourist reviews, 63% positive and 10.49% negative evaluations were made about the protection of the region. In addition, the awareness of the tourists about whether the locations they visit are World Heritage Sites was 30.6%. Tourists explained their negative opinions about the locations they visited with the words "extra_payment", "crowded", and "steep". Another finding of the study was that the area is adequately protected. There were, however, some concerns related to protection. The most striking aspects of protection concerns were those in which the words "painted" and "drawn" were used, an important indicator of visitor sensitivity to the protection of frescoes in rock churches. Since the destruction of natural and cultural assets in the region is unacceptable, recommendations are made to take the necessary measures to prevent such damage.

DC Universe is a fictional universe in which a collection of superheroes and super villains based on characters that appear in comic books by DC Comics is in it. DC Comics itself is the largest and oldest comic book publisher that produces and displays superheroes and super villains. To start a super hero-themed business, there are a number of business examples that can be used as references and can also be used to reap profits from the business, namely, rental of comics, selling merchandise, making clothing lines, making cosplay costumes, making superhero-themed foods, selling action figure. In an effort to start a super hero-themed business, especially the DC Universe theme, it is necessary to pay attention / listen to DC Universe consumers in Indonesia. Classification of opinions or sentiment analysis is one way to find out about a person or group of people towards certain products, services, issues or groups from various social media platforms and the internet. Twitter is one of the social media that is loved by the people of Indonesia. This research tries to utilize what was written by Twitter social media users or better known as a tweet. Tweets will be processed by text mining and processed again using the Naïve Bayes Classifier algorithm.

In this paper, we propose a new approach for evaluating research projects and programs. According our approach, the improvement might be achieved by adopting a results-based and a project portfolio approach, and assuring a research and technology development (RTD) indicators documentation through a standard and comprehensive indicator description, named indicator template. The results-based approach will assure a consistent indicators structure, according to the results chains and a strong connection between ex ante and ex post impact evaluation. The project portfolio approach will assure a tight integration of the research performance indicators, especially between policies goals and program results. And, finally defining a comprehensive indicator template it will be possible to understand better the indicators, develop a detailed analysis, based on the business intelligence techniques, such as OLAP (On-Line Analytical Processing), data mining and text mining. According our knowled...

Named Entities Recognition (NER) has become one of the major issues in Information Retrieval (IR), knowledge extraction, and document classification. This paper addresses a particular case of NER, acronym expansion (or definition) when this expansion does not exist in the document using the acronym. Since acronyms may obviously expand into several distinct sets of words, this paper provides nine quality measures of the relevant definition prediction based on mutual information (MI), cubic MI (MI3), and Dice's coefficient. A combinaison of these statistical measures with the cosine approach is proposed. Experiments have been run on biomedical domain where acronyms are numerous. The results on our biomedical corpus showed that the proposed measures were accurate devices to predict relevant definitions. Povzetek: Predstavljene so metode spletnega preiskovanja dvoumnih akronimov v domeni biomedicinskih baz.

This report provides a description of the Open Boek intelligent retrieval system version 3, and of its care and feeding. It combines the user manual and the administration guide. Finally, it provides detailed descriptions of the scripts and fileformats.

Enhancing the educational corporations is truly challenging mission due to the highly competitive nature of the business. Currently, there is emerging development within organizations to capitalize on their internal resources. This paper is taking data mining approach to strategize marketing for postgraduate studies by means of cluster analysis. The experiments were carried out using Oracle Data Miner tool, results are analyzed and discussed.

Bibliometric studies have the particularity of analyzing the performance of a research field/area. The present study, following this premise, will map the works of a Postgraduate program at a Brazilian Federal University. For this purpose, text mining and clustering techniques are used to analyze 609 works, including theses and dissertations. The use of text mining techniques and visualization of similarities provided the indication of groupings of knowledge areas. The similarity analysis indicates the little interaction between the fields of knowledge, characterizing the existence of “conceptual islands”. This result may have been impacted by the way the data was extracted in the similarity analysis. The analysis also made it possible to verify which themes arouse greater interest in the program in relation to the research being carried out.

This paper introduces the concept of a classification tool for Web pages called WebClassify, which uses modified naive Bayesian algorithm with multinomial model to classify pages into various categories. The tool starts the classification from downloading training Web text from Internet, preparing the hypertext for mining, and then storing Web data in a local database. The paper also gives an

1 (English) The spread of English as the Lingua Franca of international communication has given rise to meaningful language contact phenomena in the world's languages like loanwords and pseudo-loanwords, namely, words from one language (the donor language) are adopted by another language (the recipient language) sometimes becoming naturalized (Gusmani 1973). From this perspective, it is thus interesting to observe their behaviour in real language use. In particular, this study investigates Anglicisms and pseudo-Anglicisms found in the newspaper discourse of Brexit by way of a bilingual corpus collected from two Italian newspapers, i.e. La Repubblica and Il Corriere della Sera and two British newspapers, i.e. The Independent and The Guardian selected for both their authoritativeness and their extensive readership. The exit of the United Kingdom from the European Union was chosen because it is a widely covered topic both in the Italian and in the British press, thus providing abundant material for comparative analysis, as well as offering useful data in order to explore linguistic variation. It was useful for building an electronic corpus which was retrieved from the digital archives of the newspapers' websites in order to carry out an automated text analysis. The corpus includes articles collected during the periods that both preceded and followed the Brexit referendum. In order to carry out the analysis, corpus-driven methodology was used, namely an approach that lets hypotheses emerge from corpus observation (Tognini-Bonelli 2001). The investigation was carried out by way of the software TalTac2, and the automated text analysis, as a result, turned out to be invaluable in order to investigate and monitor the newspapers' vocabulary which included technical terms from the fields of politics, economics and finance as well as general language words. In order to design and sample a representative corpus, the parameters proposed by Biber (1993) were used to identify descriptive criteria so as to select and balance the population. The aim of this study is to get an overview of the Brexit discourse as used in the two countries' newspapers' vocabulary and terminology (of the two countries) by using text mining to compare and categorize the whole corpus as a collection of texts and, then, to cluster documents on the basis of the lexical similarity of the vocabulary to establish semantic fields or conceptual areas. Furthermore, by way of the lexical and textual analysis, this study also investigates Anglicisms and pseudo-Anglicisms in the Italian newspapers, identifying and analyzing a list of English words used in Italian. The two British newspapers serve as a reference corpus to compare to the list of Anglicisms extracted from the Italian corpus. The articles retrieved from the British newspapers serve to find out which words are typical of each corpus and to identify pseudo-anglicisms, namely new words that seem to be English forms, even though they do not exist in English, or if they do exist, they have a clearly different meaning. Lastly, the data gathered from the bilingual corpus analysis were later compared with other wider corpora included in SketchEngine and on the Brigham Young University platform in order to make generalizations about the distribution of Anglicisms and pseudo-Anglicisms in general language corpora. Abstract 2 (Italian) La diffusione e l'affermazione dell'inglese come lingua franca della comunicazione internazionale ha generato fenomeni significativi di contatto linguistico come i prestiti e i falsi prestiti, ossia parole originariamente nate in una lingua modello che entrano a far parte di un'altra lingua (lingua replica) alla quale vengono talvolta assimilate e adattate (Gusmani 1973). È quindi interessante osservarne l'uso e l'andamento in testi autentici che presentano la lingua nel suo uso corrente. Questo studio analizza gli anglicismi e i falsi anglicismi nel discorso giornalistico della Brexit, attraverso un corpus tratto dai quotidiani italiani La Repubblica e Il Corriere della Sera e dai quotidiani britannici The Guardian e The Independent, che sono stati selezionati per la loro diffusione 2 VALERIA FIASCO JADT 2018 : 14 es Journées internationales d'Analyse statistique des Données Textuelles e la loro autorevolezza. La scelta della tematica dell'uscita del Regno Unito dall'Unione Europea è stata dettata da diversi fattori, tra i quali l'ampia diffusione dell'argomento nella stampa italiana e in quella britannica, dando la possibilità di creare un corpus per realizzare un'analisi comparativa attraverso l'esplorazione della variazione linguistica. Dal momento che queste riviste offrono una versione online che mette a disposizione un archivio digitale consultabile, sono particolarmente adatte per creare un corpus che può essere esaminato attraverso l'analisi automatica del testo. Il corpus è composto da articoli raccolti durante il periodo che precede e segue il referendum della Brexit e la metodologia utilizzata per condurre l'analisi è di tipo corpus-driven, ossia un approccio esplorativo in cui, partendo dall'osservazione del corpus, si arriva alla formulazione delle ipotesi (Tognini-Bonelli 2001). Il software TalTac2 e l'analisi automatica dei testi sono stati estremamente preziosi per esaminare e monitorare il lessico della stampa che include termini tecnici della politica, dell'economia e della finanza, insieme a parole che fanno parte del lessico comune. Per progettare il corpus, sono stati utilizzati i parametri proposti da Biber (1993) con lo scopo di identificare i criteri descrittivi per selezionare e bilanciare la popolazione all'interno del corpus. L'obiettivo di questa ricerca è offrire un'analisi del lessico e della terminologia utilizzata nel discorso sulla Brexit nei quotidiani italiani e inglesi attraverso il text mining per raffrontare i testi che compongono il corpus, categorizzarli e raggrupparli sulla base di somiglianze lessicali per individuare i campi semantici e le aree concettuali. Inoltre, l'analisi lessicale e testuale ha consentito l'identificazione degli anglicismi e dei falsi anglicismi nei quotidiani italiani, mentre il corpus dei quotidiani britannici ha svolto la funzione di corpus di riferimento per paragonare la lista degli anglicismi estratta dal corpus italiano con i dati raccolti nel corpus britannico, capire quali parole sono tipiche di ogni lingua e identificare i falsi anglicismi, vale a dire parole che presentano una forma inglese, che però non esistono nel vocabolario originario o nel caso in cui esistano, il loro significato è completamente differente. Infine, i dati raccolti dall'analisi del corpus bilingue sono stati successivamente confrontati con altri corpora più ampi, consultabili su SketchEngine e sulla piattaforma della Brigham Young University con lo scopo di fare delle generalizzazioni sulla distribuzione degli anglicismi e dei falsi anglicismi in corpora non specialistici.

In this paper, we establish Fog Index (FI) as a text filter to locate the sentences in texts that contain connected biomedical concepts of interest. To do so, we have used 24 random papers each containing four pairs of connected concepts. For each pair, we categorize sentences based on whether they contain both, any or none of the concepts. We then use FI to measure difficulty of the sentences of each category and find that sentences containing both of the concepts have low readability. We rank sentences of a text according to their FI and select 30 percent of the most difficult sentences. We use an association matrix to track the most frequent pairs of concepts in them. This matrix reports that the first filter produces some pairs that hold almost no connections. To remove these unwanted pairs, we use the Equally Weighted Harmonic Mean of their Positive Predictive Value (PPV) and Sensitivity as a second filter. Experimental results demonstrate the effectiveness of our method.

Social media data is increasingly used to gain insights into trends in mental health, but prior studies aimed at confirming a link between online expression of suicidal ideation on social media and actual suicide deaths have been inconclusive. Using comprehensive six-year data sets of Twitter posts and suicide deaths in Japan, we examine the diurnal relationship between the proportional incidence of a suicide-related keyword, “kietai” (“I want to disappear”), and suicide deaths with an OLS regression model. We also use co-occurrence analysis to reveal changes in the linguistic context of the suicide-related keyword at different hours of the day. We find a clear diurnal pattern in the use of this suicide-related keyword, peaking between 1am and 5am. This diurnal trend is positively correlated with suicide deaths among younger cohorts (ages 15 to 44), but the correlation is negative among older adults (45 and over). The correlation among young adults strengthens when a delay between tweet incidence and suicide deaths is included. Compared to daytime tweets, nighttime tweets exhibited a stronger relationship between words related to self-disgust and words directly indicating suicidal intent. This study confirms the hypothesised link between online suicidal ideation and suicide death. Despite frequent flippant use of the keyword, the consistent correlation and the diurnal changes in the context of the keyword's usage demonstrate the value of social media data to the study of mental health trends in groups at risk of suicide.

This study used the data of questionnaires conducted by a local TV program on child rearing to analyze the worries and needs of parents raising children in Yamanashi prefecture. Data from questionnaires conducted by the program for every year between 2007 and 2014 were used, with a total of 1962 respondents. Results from a multiple-choice question revealed that a majority of respondents had worries regarding disciplining their children. Text mining analyses of free-answer questions indicated that respondents are unconfident regarding their methods of reprimanding their children and want to know better ways to respond when their children engage in undesirable behavior. Text mining also revealed that there is strong demand among child rearing parents, mostly mothers, on information regarding child-friendly places to go out to.

This paper describes a text mining tool that performs two tasks, namely document clustering and text summarization. These tasks have, of course, their corresponding counterpart in "conventional" data mining. However, the textual, unstructured nature of documents makes these two text mining tasks considerably more difficult t han their data mining counterparts. In our system document clustering is performed by using

Most of company product or service depends on user experience. So finding feedback of each user is tough but social platform reduce this dependency as people provide this information in form of comments. This paper has work in this domain of sentiment mining where user product based comments were studied for identifying its experience. Paper has extract the phrase feature from the comments dataset. Phrase were arrange in graph where nodes were phrase and edge have similarity sentiment count used in comments. Each phrase was further processed by frog Jump genetic algorithm for clustering of phrases as per sentiment class. In frog jump clustering phrase graph structure was used as fitness value evaluation. Paper has found that memeplex operation of frog algorithm has increase the clustering efficiency of work. Real twitter dataset was used for the comment sentiment analysis. Result shows that Phrase based Frog Jump sentiment detection has increase the accuracy, precision of work as compared to existing algorithm.

Com o crescimento da Educação a distância (EAD), os fóruns de discussão online se apresentam como um importante instrumento no processo de ensino-aprendizagem. No entanto, à medida que as discussões em fóruns crescem torna-se difícil para professores/tutores acompanhar as discussões e consequentemente avaliar as contribuições dos estudantes. Diante disso, este artigo avalia um modelo de avaliação automática de postagens em um fórum educacional conduzido em língua portuguesa por meio de 4 classificadores. Os resultados obtidos demonstram o bom desempenho do multilayer perceptron (MLP) para classificação das postagens e evidencia a possibilidade do modelo avaliado ser utilizado para fins educacionais.

An important task in text mining is finding the dominant topics (and their associated documents) in a collection of documents. While traditional clustering techniques, e.g., hierarchical clustering and K-means, are often used for this task, this paper explores a new clustering algorithm which is based on a shared nearest neighbor approach. Unlike traditional clustering algorithms, not all the data is

Beta endorphin is a neurotransmitter and is involved in functions like enhancement of immune system, deceleration of cancer cell growth and induction of euphoria and relaxation. It is an opioid like neuropeptide synthesized in neurons of hypothalamus and pituitary gland. The study of its structure and its interaction with opioid receptors can throw light on its neuropsychopharmacology. Protein data bank does not contain the structural information of beta endorphin and mu opioid receptor. Hence, in the present study, we aimed at predicting their three dimensional structures. Owing to homologues with low sequence identity and unsatisfactory results from threading methods, we resorted to ab initio modelling. Quark algorithm was used for beta endorphin structure prediction. The structure of mu opioid receptor was modelled by I TASSER simulations. The quality of the protein models were evaluated with PROCHECK server. Functionally important regions in beta endorphin were located using ConSurf web server. Docking studies were performed on beta endorphin with three opioid receptors mu, kappa and gamma to anticipate residues important for binding.

Se presenta una propuesta de modelo para el diseño de estrategias de aprendizaje móvil, que fue realizado en 2012 por investigadores del grupo Proventus de la universidad de la Sabana, en el marco de un proyecto de investigación que tiene como objetivo el diseño de estrategias de aprendizaje móvil, para el desarrollo de competencias en manejo de información. El modelo ha sido diseñado como respuesta a la falta de un modelo que pudiera ser utilizado por un docente para diseñar sus propias estrategias de aprendizaje móvil y no supiera por donde comenzar. Para su diseño se realizo una revisión inicial sobre los diferentes modelos que existen en la actualidad para el diseño de estrategias de aprendizaje móvil y se tomo como base el proceso que siguen los investigadores para diseñar sus estrategias de aprendizaje en el programa de maestría en informática educativa de la Universidad de la Sabana.

In finance task domain, it is indispensable to get and analyze information as quickly as possible. Analyst's reports are one of the important information in asset management, and these include a large amount of text information. However, it is very difficult to handle text information of analyst's reports, few research and development have been conducted. In [5] and [6] we explored the feasibility to extract valuable knowledge for asset management through text mining using analyst's reports as text data. And we found the effectiveness of keyword information. In this paper we make further research of analyst's reports. From empirical study on the practical data, we have confirmed the effectiveness of using keyword information and numerical information together: (1) the effectiveness of keyword information is different by the direction of change of earning estimate; (2) the keyword of "Upward (or Downward) surprise in forecast" has strong effect to stock price return.

BioCaster is an ontology-based text mining system for detecting and tracking the distribution of infectious disease outbreaks from linguistic signals on the Web. The system continuously analyzes documents reported from over 1700 RSS feeds, classifies them for topical relevance and plots them onto a Google map using geocoded information. The background knowledge for bridging the gap between Layman's terms and formal-coding systems is contained in the freely available BioCaster ontology which includes information in eight languages focused on the epidemiological role of pathogens as well as geographical locations with their latitudes/longitudes. The system consists of four main stages: topic classification, named entity recognition (NER), disease/location detection and event recognition. Higher order event analysis is used to detect more precisely specified warning signals that can then be notified to registered users via email alerts. Evaluation of the system for topic recognition and entity identification is conducted on a gold standard corpus of annotated news articles.

The experience of the Golf Practice Class with Lodging and the Change of Life Skills through it

The purpose of this study is to examine the experience of the golf practice class with lodging and the longitudinal change of life skills through it. The subjects of 11 university students (male=9, female=2, mean age = 19.30, SD = 1.17) were asked to answer a questionnaire that was composed of socio-demographic questions, the Experience Scale in University Physical Education Classes (ESUPEC), impressions of golf practice class, and Life Skills Scale for University Students (LSSUS). Descriptive statistics and text mining were conducted to reveal the experience of the golf practice class. Paired t-tests and calculation of Cohen’s d were conducted to examine longitudinal change of life skills. The results of descriptive statistics showed that mean scores of ESUPEC were high compared with a previous research. The results of text mining showed that the words of “enjoy” and “difficult” appeared frequently. In addition, some word networks such as including positive psychological aspects were shown. Paired t-tests revealed that the scores of “Leadership”, “Scheduling” and “Self-Esteem” of LSSUS increased significantly, their Cohen’s d ranged from 0.43 to 0.63. The results above indicate that sports activities with lodging might be more effective to obtain psychosocial skills.

This study empirically evaluates the effectiveness of different feature types for the classification of the first language of an author. In particular, it examines the utility of psycholinguistic features, extracted by the Linguistic Inquiry and Word Count (LIWC) tool, that have not previously been applied to the task of author profiling. As LIWC is a tool that has been developed in the psycholinguistic field rather than the computational linguistics field, it was hypothesized that it would be effective, both as a single type feature set because of its psycholinguistic basis, and in combination with other feature sets, because it should be sufficiently different to add insight rather than redundancy. It was found that LIWC features were competitive with previously used feature types in identifying the first language of an author, and that combined feature sets including LIWC features consistently showed better accuracy rates and average F measures than were achieved by the same feature sets without the LIWC features. As a secondary issue, this study also examined how effectively first language classification scaled up to a larger number of possible languages. It was found that the classification scheme scaled up effectively to the entire 16 language collection from the International Corpus of Learner English, when compared with results achieved on just 5 languages in previous research.

Analyzing stock market trends and sentiment is an interdisciplinary area of research being undertaken by many disciplines such as Finance, Computer Science, Statistics, and Economics. It has been well established that real time news plays a strong role in the movement of stock prices. With the advent of electronic and online news sources, analysts have to deal with enormous amounts of real-time, unstructured streaming data. In this paper, we present an automated text mining based approach to aggregate news stories from diverse sources and create a News Corpus. The Corpus is filtered down to relevant sentences and analyzed using Natural Language Processing (NLP) techniques. A sentiment metric, called NewsSentiment, utilizing the count of positive and negative polarity words is proposed as a measure of the sentiment of the overall news corpus. We have used various open source packages and tools to develop the news collection and aggregation engine as well as the sentiment evaluation engine. Extensive experimentation has been done using news stories about various stocks. The time variation of NewsSentiment shows a very strong correlation with the actual stock price movement. Our proposed metric has many applications in analyzing current news stories and predicting stock trends for specific companies and sectors of the economy.

This paper describes a text mining tool that performs two tasks, namely document clustering and text summarization. These tasks have, of course, their corresponding counterpart in "conventional" data mining. However, the textual, unstructured nature of documents makes these two text mining tasks considerably more difficult than their data mining counterparts. In our system document clustering is performed by using the Autoclass data mining algorithm. Our text summarization algorithm is based on computing the value of a TF-ISF (term frequency -inverse sentence frequency) measure for each word, which is an adaptation of the conventional TF-IDF (term frequency -inverse document frequency) measure of information retrieval. Sentences with high values of TF-ISF are selected to produce a summary of the source text. The system has been evaluated on real-world documents, and the results are satisfactory.

A fully automated, self-driving car can perceive its environment, determine the optimal route, and drive unaided by human intervention for the entire journey. Connected autonomous vehicles (CAVs) have the potential to drastically reduce accidents, travel time, and the environmental impact of road travel. Such technology includes the use of several sensors, various algorithms, interconnected network connections, and multiple auxiliary systems. CAVs have been subjected to attacks by malicious users to gain/deny control of one or more of its various systems. Data security and data privacy is one such area of CAVs that has been targeted via different types of attacks. The scope of this study is to present a good background knowledge of issues pertaining to different attacks in the context of data security and privacy, as well present a detailed review and analysis of eight very recent studies on the broad topic of security and privacy related attacks. Methodologies including Blockchain, Named Data Networking, Intrusion Detection System, Cognitive Engine, Adversarial Objects, and others have been investigated in the literature and problem-and context-specific models have been proposed by their respective authors.

Metric and kernel learning are important in several machine learning applications. However, most existing metric learning algorithms are limited to learning metrics over low-dimensional data, while existing kernel learning algorithms are often limited to the transductive setting and do not generalize to new data points. In this paper, we study metric learning as a problem of learning a linear transformation of the input data. We show that for high-dimensional data, a particular framework for learning a linear transformation of the data based on the LogDet divergence can be efficiently kernelized to learn a metric (or equivalently, a kernel function) over an arbitrarily high dimensional space. We further demonstrate that a wide class of convex loss functions for learning linear transformations can similarly be kernelized, thereby considerably expanding the potential applications of metric learning. We demonstrate our learning approach by applying it to large-scale real world problems in computer vision and text mining.

User satisfaction from education institution, health service or other bussiness are often measured using questionnaire with scoring grade system. Generally many comments, advices and suggestions from questionnaire complements is not processed systematically due to lack of method to process. whereas such information might be very useful as additional source in decision making. Opinion mining and sentiment analysis are well suited to address the issue. The objective of this study is to develop sentiment classifacation system using Naive Bayes Classifier (NBC) and Support Vector Machine (SVM). Opinion to be classified was Indonesian textual comments from academic questionnaire. Data contained of 1000 comments that was sampled as multi domain opinion, represented many objects such as lecturer, class room, etc. Features used for classification was selected from word and phrasa using in the comments. The weighting scheme that we used was TF/IDF. The results show that SVM gives good performance as compared with NBC.

The Indonesian tourism industry continues to develop and has become the core of the nation economy. Indonesia is known for its wealth of natural beauty, which can be used as a potential for tourism business. Garut is a city in Indonesia known for its beach, mountains, and culinary arts. The purpose of this study is to create a priority map of tourist attractions that can be utilized by local governments. Sentiment analysis was used 413,175 netizen comments via the social media platforms Instagram and Google reviews. Data was collected from January 2018-February 2019. The results show that the number of positive comments is significantly higher than the number of negative comments. Beach tourism a serious priority; not only is it the most preferred tourist attraction, but also the type that gets the most negative comments. While the main problem for Garut Regency tourism is hygiene, garbage is either overlapping or scattered, preventing Garut from having all-around tourism charm instead of being superior only in the sector of natural beauty. The suggestions in this research can be used as proposals for improving and developing tourism to realize a dignified 'Garut Charm'.

In the present state of digital world, computer machine do not understand the human's ordinary language. This is the great barrier between humans and digital systems. Hence, researchers found an advanced technology that provides information to the users from the digital machine. However, natural language processing (i.e. NLP) is a branch of AI that has significant implication on the ways that computer machine and humans can interact. NLP has become an essential technology in bridging the communication gap between humans and digital data. Thus, this study provides the necessity of the NLP in the current computing world along with different approaches and their applications. It also, highlights the key challenges in the development of new NLP model.

Text mining is the process of discovering information in large text collections, and automatically identifying interesting patterns and relationships in textual data. It is a relatively new research area, which has recently raised much interest among the research and industry communities, mainly due to the continuously increasing amount of information available on the Web and elsewhere. Text mining is a highly interdisciplinary research area, bringing together research insights from the fields of data mining, natural language processing, machine learning, and information retrieval. In particular, text mining is closely related to the older area of data mining, which targets the extraction of interesting information from data records, although text mining is allegedly more difficult, as the source data consists of unstructured collections of documents rather than structured databases.