News Text Data Research Articles

Text Spillover: Measuring connectedness of financial institutions based on news text data

The interconnectedness of financial institutions can facilitate the contagion of financial crises and poses significant risks to financial stability. While well-established econometric spillover measures exist, they are usually based on numerical data, which may not always be available or timely. This paper explores the feasibility of using text data analysis as an alternative measure of interconnectedness between financial institutions. A novel news-based spillover measure is constructed and applied to a system of five banks, using an up-to-date dataset of nearly 47,000 news articles. The results of this text spillover index are compared to the Diebold-Yilmaz spillover index for validation. The findings indicate that text data analysis can provide valuable insight into the interconnectedness of financial institutions and substantiate the underlying hypothesis that the co-occurrence of actors in news articles allows direct conclusions about their connectedness. This enables new types of analysis and extends spillover measures to a wider range of actors

Indonesian Fake News Classification Using Transfer Learning in CNN and LSTM

Fake news spreads quickly and is challenging to stop due to the ease of accessing and sharing information online. Deep learning techniques are a method that can be used to identify fake news quickly and accurately. The types of neural networks commonly utilized in deep learning architectures include Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM), which can perform well when managing the task of classifying fake news, according to several pertinent studies. Regarding handling instances of Indonesian fake news classification, this study compares how well the CNN and LSTM models perform. However, given that Indonesian is a low-resource language with scant documentation, it is challenging to build an adequate data set. At the same time, the CNN and LSTM classification models require significant training data. We proposed a transfer learning method by combining two classification models with a pre-trained IndoBERT language model. 1340 news text data were used, including 643 actual news texts from CNN Indonesia, Liputan6, and Detik and 697 fake news texts from TurnBackHoax. As a result, the performance of the combination of the LSTM classification model with IndoBERT outperformed that of the CNN classification model with IndoBERT, which only produced an accuracy of 92.91%, down by 6%, and was able to produce an accuracy of up to 97.76%, an increase of 4.8% from before. Furthermore, the results show that the LSTM classification model outperforms the CNN classification model in capturing the representation created by IndoBERT. Additionally, these insights may serve as a basis for future research on identifying fake news in Indonesia, helping to improve methods for combatting misinformation in Indonesia.

Real-Time Extraction of News Events Based on BERT Model

Abstract For the large number of news reports generated every day, in order to obtain effective information from these unstructured news text data more efficiently. In this paper, we study the real-time crawling of news data from news websites through crawling techniques and propose a BERT model-based approach to extract events from news long text. In this study, NetEase news website is selected as an example for real-time extraction to crawl the news data of this website. BERT model as a pre-trained model based on two-way encoded representation of transformer performs well on natural language understanding and natural language generation tasks. In this study, we will fine-tune the training based on BERT model on news corpus related dataset and perform sequence annotation through CRF layer to finally complete the event extraction task. In this paper, the DUEE dataset is chosen to train the model, and the experiments show that the overall performance of the BERT model is better than other network models. Finally, the model of this paper is further optimised, using the ALBERT and RoBERTa models improved on the basis of the BERT model, experiments were conducted, the results show that both models are improved compared to the BERT model, the ALBERT model algorithm performs the best, the model algorithm's F1 value is 1% higher than that of BERT. The results show that the performance is optimised.

Which news topics drive economic prosperity in China?

Precise and real-time measurements of economic prosperity are vital to a country's economic system. This study aims to identify news topics that promoted economic prosperity in China from 2011-2021. By extracting economic topics from news text data, we construct a news coincidence index with comprehensive information and strong timeliness and reveal the trend of topic contribution. The Latent Dirichlet Allocation (LDA) topic model is applied to extract economic topics from the news. We use a mixed-frequency dynamic factor model to track rapid economic development without using high-frequency weekly and daily data. We identify the six most influential topics and investigate their evolution, which may serve as a reference for economic construction and regulation.

Examining semantic (dis)similarity in news through news organizations’ ideological similarity, similarity in truthfulness, and public engagement on social media: a network approach

Abstract The rise of homogenization and polarization in the news may inhibit individuals’ understanding of an issue and the functioning of a democratic society. This study applies a network approach to understanding patterns of semantic similarity and divergence across news coverage. Specifically, we focus on how (a) inter-organizational networks based on media ideology, (b) inter-organizational networks based on news truthfulness, and (c) public engagement that news articles received on social media may affect semantic similarity in the news. We use large-scale user logs data on social media platforms (i.e., Facebook and Twitter) and news text data from more than 100 news organizations over 10 months to examine the three potential processes. Our results show that the similarity between news organizations in terms of media ideology and news truthfulness is positively associated with semantic similarity, whereas the public engagement that news articles received on social media is negatively associated with semantic similarity. Our study contributes to theory development in mass communication by shifting to a network paradigm that connects news organizations, news content, and news audiences. We demonstrate how scholars across communication disciplines may collaborate to integrate distinct theories, connect multiple levels, and link otherwise separate dimensions. Methodologically, we demonstrate how synchronizing network science with natural language processing and combining social media log data with text data can help to answer research questions that communication scholars are interested in. The findings’ implications for news polarization are discussed.

뉴스 텍스트를 활용한 국내 자동차 시장의 브랜드 속성 기반 시장 구조 분석

We propose an approach to explore the discourse generated by the public press to draw meaningful insights on brands, brand attributes, and the brand attributes-based market structure. Focusing on the Korean automobile market, we collect 346,795 news articles related to the automobile industry. Then, we employ a machine learning-based text mining model to extract the semantic structure of brands and brand attributes reflected in the news text data. Specifically, we utilize the Word2vec model to assign each word as a vector, and measure the semantic (dis)similarity as the vector distance between word vectors. This approach enables us to find key brand attributes in the Korean automobile market, such as “quality”, “luxury”, and “trust”, and to measure the strength of their associations with automobile brands. Using the associations between brand and brand attributes, we visualize the brand-level market structure via perceptual maps that reveal the competitive nature of the market at the brand level. Based on these results, we summarize meaningful insights for the brand management and marketing strategy.

Online Newspaper Clustering in Aceh using the Agglomerative Hierarchical Clustering Method

The rapid progress in the field of information technology, especially the internet, has given birth to a lot of information. The ease of publishing an article on a website causes an explosion of news pages which will certainly confuse readers. The diversity and the increasing number of news articles make it increasingly difficult for internet users to find news and large piles of news data on online newspaper sites in Aceh. The grouping of text documents is needed to classify news in online newspapers in Aceh based on the content contained in news articles. In this study, the process of grouping online news in Aceh was tried using the Agglomerative Hierarchical Clustering method. News is grouped with a Bottom-Up design strategy that starts with placing each object as a cluster then combined into a larger cluster based on the similarity of keywords in each news, then the cluster results are compared and put into each news category. The research design was carried out in a structured manner using data flow diagrams in forming the research framework. The study was conducted by taking online news text data on 10 online news websites in Aceh from July 2016 to March 2017 with 1000 randomly generated documents. The process of crawling news data is done using a php script which will only take text files from the news on the website. News grouping is done based on religion, politics, law, sports, tourism, education, culture, economy and technology. The results of the grouping performance of the Agglomerative Hierarchical Clustering method in this study have an average accuracy of 89.84%.

Sentiment Classification of News Text Data Using Intelligent Model

Text sentiment classification is a fundamental sub-area in natural language processing. The sentiment classification algorithm is highly domain-dependent. For example, the phrase “traffic jam” expresses negative sentiment in the sentence “I was stuck in a traffic jam on the elevated for 2 h.” But in the domain of transportation, the phrase “traffic jam” in the sentence “Bread and water are essential terms in traffic jams” is without any sentiment. The most common method is to use the domain-specific data samples to classify the text in this domain. However, text sentiment analysis based on machine learning relies on sufficient labeled training data. Aiming at the problem of sentiment classification of news text data with insufficient label news data and the domain adaptation of text sentiment classifiers, an intelligent model, i.e., transfer learning discriminative dictionary learning algorithm (TLDDL) is proposed for cross-domain text sentiment classification. Based on the framework of dictionary learning, the samples from the different domains are projected into a subspace, and a domain-invariant dictionary is built to connect two different domains. To improve the discriminative performance of the proposed algorithm, the discrimination information preserved term and principal component analysis (PCA) term are combined into the objective function. The experiments are performed on three public text datasets. The experimental results show that the proposed algorithm improves the sentiment classification performance of texts in the target domain.

Applications of Deep Learning in News Text Classification

The advancement in technology is taking place with an accelerating pace across the globe. With the increasing expansion and technological advancement, a vast volume of text data are generated everyday, in the form of social media platform, websites, company data, healthcare data, and news. Indeed, it is a difficult task to extract intriguing patterns from the text data, such as opinions, summaries, and facts, having varying length. Because of the problems of the length of text data and the difficulty of feature value extraction in news, this paper proposes a news text classification method based on the combination of deep learning (DL) algorithms. In order to classify the text data, the earlier approaches use a single word vector to express text information and only the information of the relationship between words were considered, but the relationship between words and categories was ignored which indeed is an important factor for the classification of news text. This paper follows the idea of a customized algorithm which is the combination of DL algorithms such as CNN, LSTM, and MLP and proposes a customized DCLSTM-MLP model for the classification of news text data. The proposed model is expressed in parallel with word vector and word dispersion. The relationship among words is represented by the word vector as an input of the CNN module, and the relationship between words and categories is represented by a discrete vector as an input of the MLP module in order to realize comprehensive learning of spatial feature information, time-series feature information, and relationship between words and categories of news text. To check the stability and performance of the proposed method, multiple experiments were performed. The experimental results showed that the proposed method solves the problems of text length, difficulty of feature extraction in the news text, and classification of news text in an effective way and attained better accuracy, recall rate, and comprehensive value as compared to the other models.

A Big Data Analysis of Job Position Status Based on Natural Language Processing

With the development of the Internet and computer technology, news on the Internet has gradually been able to influence the development of the economy. In some special periods, with the development of major events, the economy and finance have undergone similar development and changes. This paper proposes a simplified natural language processing method for multi-grained text. First, the news text data is classified according to sentiment, and the correlation between corporate loans and job retention ability is further analysed. The experimental results show that under certain conditions, the Loans issued to SMEs are more conducive to retaining the number of jobs than large enterprises.

Research on News Text Classification Based on Deep Learning Convolutional Neural Network

Aiming at the problems of low classification accuracy and low efficiency of existing news text classification methods, a new method of news text classification based on deep learning convolutional neural network is proposed. Determine the weight of the news text data through the VSM (Viable System Model) vector space model, calculate the information gain of mutual information, and determine the characteristics of the news text data; on this basis, use the hash algorithm to encode the news text data to calculate any news. The spacing between the text data realizes the feature preprocessing of the news text data; this article analyzes the basic structure of the deep learning convolutional neural network, uses the convolutional layer in the convolutional neural network to determine the change value of the convolution kernel, trains the news text data, builds a news text classifier of deep learning convolutional neural network, and completes news text classification. The experimental results show that the deep learning convolutional neural network can improve the accuracy and speed of news text classification, which is feasible.

Using Computational Text Analysis Tools to Study African Online News Content

After radio and television, online media are fast becoming a primary source of information for many Africans. With this increase, it is becoming necessary for media researchers to explore ways to better understand production, content and reception patterns of online news in the continent. This paper introduces freely available tools for systematic and (semi-)automated collection, storage and analysis of digital news that builds on recent advances in the computational power of personal computers, and the decreasing costs of storing large amounts of data. I start by describing existing challenges in the collection of online news text data, including the limited amount of African news content in commercial databases, and the methodological shortcomings of using commercial search engines. Then, I present a four-stage approach using packages written in the open-source R programming language to automate the collection of online news content (web scraping); transform this content for easier storage and analysis (data processing); use computational text analysis tools to describe and categorise data; and present the results in ways that are easier to understand (data visualisation). The paper concludes with a summary of recommendations for using computational methods to study African communication phenomena.

Indonesian news classification using convolutional neural network

<span>Every language has unique characteristics, structures, and grammar. Thus, different styles will have different processes and result in processed in Natural Language Processing (NLP) research area. In the current NLP research area, Data Mining (DM) or Machine Learning (ML) technique is popular, especially for Deep Learning (DL) method. This research aims to classify text data in the Indonesian language using Convolutional Neural Network (CNN) as one of the DL algorithms. The CNN algorithm used modified following the Indonesian language characteristics. Thereby, in the text pre-processing phase, stopword removal and stemming are particularly suitable for the Indonesian language. The experiment conducted using 472 Indonesian News text data from various sources with four categories: ‘hiburan’ (entertainment), ‘olahraga’ (sport), ‘tajuk utama’ (headline news), and ‘teknologi’ (technology). Based on the experiment and evaluation using 377 training data and 95 testing data, producing five models with ten epoch for each model, CNN has the best percentage of accuracy around 90,74% and loss value around 29,05% for 300 hidden layers in classifying the Indonesian News data.</span>

Earthquake Knowledge Graph Constructing Based on Social Intercourse Using BiLSTM-CRF

Nowadays, the earthquake has become a very serious topic. The earthquake-related information always appears first in social media. Constructing an earthquake knowledge graph can help dealing with earthquake news social media text data. This paper proposed the BiLSTM-CRF (Bi-directional Long Short Term Memory-Conditional Random Field) model to construct earthquake knowledge graph using news text data. The BiLSTM-CRF model identified the entities and then writes entities and their types to the table so that the entities and relationships between entities can be extracted for earthquake. The entities and relationships between entities have been combined with the form of RDF (Resource Description Framework) to construct earthquake knowledge graph on Neo4j database.

Impact of News on the Trend of Stock Price Change: an Analysis based on the Deep Bidirectiona LSTM Model

Abstract With the in-depth study of the stock market, the impact of news media on stock prices has gradually been paid attention to. However, in previous studies, most of them used the real-time rise and fall of stock prices to reflect the impact of news on stock prices, ignoring that it takes time for investors to react to news. In this paper, considering the lag of investors’ response to stock price, we choose BIAS as a measure index after news happened for a period of time to analyze the impact of news media on stock price trends. Based on the DBLSTM (Deep Bidirectional Long Short-Term Memory) model, we establish a model to predict the short-term trend of stock prices using news text data. Experiments show that the model we adopted outperforms other models in terms of prediction accuracy.

Компьютерный анализ эмоциональной компоненты научных публикаций на примерах в физике и экономике

The results of the application of the system of automated extraction and visualization of metadata of emotionality of texts of scientific publications in the field of innovative research in physics and economics are presented. In physics, works on the direct conversion of light radiation into electrical energy are considered based on photoplasma, which is an effect of the appearance of a potential difference in the plasma of alkali metals. In this area, the stage of research work has practically been completed, and a transition to experimental design work is taking place with a subsequent assessment of the economic efficiency of the production and operation of photoconverters based on photoplasma. Such metadata is necessary for the development and implementation into practice of systems for the intellectual analysis of news text data, analytical publications to increase the efficiency of forecasting and assess development prospects in various areas, for example, in financial and economic activities to predict quotations on the securities market. The paper provides examples of sentiment analysis and its visualization of economic scientific texts and text from the field of gasdischarge plasma research. Instrumental methods of emotional analysis of texts are implemented in the R language, which allows in a short time to form the necessary packages of text analysis programs not only by professional programmers, but also by analysts.

Weekly ILI patient ratio change prediction using news articles with support vector machine

BackgroundInfluenza continues to pose a serious threat to human health worldwide. For this reason, detecting influenza infection patterns is critical. However, as the epidemic spread of influenza occurs sporadically and rapidly, it is not easy to estimate the future variance of influenza virus infection. Furthermore, accumulating influenza related data is not easy, because the type of data that is associated with influenza is very limited. For these reasons, identifying useful data and building a prediction model with these data are necessary steps toward predicting if the number of patients will increase or decrease. On the Internet, numerous press releases are published every day that reflect currently pending issues.ResultsIn this research, we collected Internet articles related to infectious diseases from the Centre for Health Protection (CHP), which is maintained the by Hong Kong Department of Health, to see if news text data could be used to predict the spread of influenza. In total, 7769 articles related to infectious diseases published from 2004 January to 2018 January were collected. We evaluated the predictive ability of article text data from the period of 2013–2018 for each of the weekly time horizons. The support vector machine (SVM) model was used for prediction in order to examine the use of information embedded in the web articles and detect the pattern of influenza spread variance. The prediction result using news text data with SVM exhibited a mean accuracy of 86.7 % on predicting whether weekly ILI patient ratio would increase or decrease, and a root mean square error of 0.611 on estimating the weekly ILI patient ratio.ConclusionsIn order to remedy the problems of conventional data, using news articles can be a suitable choice, because they can help estimate if ILI patient ratio will increase or decrease as well as how many patients will be affected, as shown in the result of research. Thus, advancements in research on using news articles for influenza prediction should continue to be pursed, as the result showed acceptable performance as compared to existing influenza prediction researches.

Research on Political Trend of North Korea Based on Big Data Text Mining Method

With the method of text mining, this paper takes the related data of Rodong Sinmun in the recent ten years as the research object, extracts the hot topics and carries on the trend analysis. With the special attribute of his speech media, this paper analyzes the political issue. By extracting nearly one million news text data, their topic content is analyzed, combining with LDA topic model, and using K-means clustering algorithm. Aiming at the limitations of the traditional K-means algorithm, it is solved on the pre-built big data analysis platform, and the structure and content of the theme are analyzed in detail. In the end, the political theme and the trend of public opinion in recent years are derived. In terms of application, it is of great significance to analyze and study Korean big data text.

High-Volume Business News: Detecting Topic-Specific Sentiments to Account for Brand Ratings and Stock Returns

Due to the highly voluminous, heterogeneous, and unstructured nature of global business news streaming at a fast pace, it has become increasingly difficult for marketing executives, corporate communications managers, and market analysts to make sense and track news media stories addressing the company and its brand. This paper responds to this challenge by developing an integrated methodology for identifying and tracking company-specific news topics as well as topic-specific sentiments from news media text data. The main contribution and novelty of the methodology lies in its ability to identify not only company-specific news topics but also to track the topic-specific sentiments (i.e. positive vs. negative tone/valence) over time. This extends prior methods and models, which have only focused on the identification of either topics or aggregate sentiments from news text data, not being able to identify and track the disaggregated topic-specific sentiments (i.e., whether the news media stories treat the firms in a positive vs. negative tone for each separate topic, over time). In addition to developing the integrated topic-sentiment model, we demonstrate the utilization value of the methodology for executives, in terms of its predictive ability for firm performance metrics. That is, we demonstrate, with a set of time-series regression models, that the topic-specific sentiments identified by the model (on Thomson Reuters Newswire data with over 300,000 news items) have the ability to predict firm performance in terms of brand ratings as well as stock returns.