Listening to the patient voice: using a sentic computing model to evaluate physicians’ healthcare service quality for strategic planning in hospitals

Online physician reviews (OPRs), also known as electronic word of mouth, have become the primary source of information for patients while making health consultation decisions. However, different techniques to analyze these reviews by machines have not been frequently applied yet in this domain. In this study, a novel method for opinion mining is being proposed to fill the existing research gap, that is, a hybrid approach to sentic computing. This approach integrates artificial intelligence and semantic web techniques to implicitly analyze OPRs in order to evaluate patient perceptions of healthcare service quality. We develop our methodology by using the following three main tasks: (1) sentence-level topic spotting (a topic-analysis procedure) to extract major topics, (2) a sentic computing framework to perform concept-level sentiment analysis (polarity detection on the categorized sentences), and (3) root cause and strengths, weaknesses, opportunities, and threats (SWOT) analyses to identify SWOT for healthcare organizations. Analyses results of 47,499 OPRs from the UK-based website (Iwantgreatcare.org) show that the proposed hybrid approach has accurately classified concept words to their corresponding topics, and it has also outperformed the similar other methods of topic extraction in the healthcare domain. The results also indicate that the proposed approach can better contribute to the overall performance analyses for healthcare organizations, which could help practitioners in improving service-quality measures based on the real voices of patients. The proposed model also provides a theoretical basis for formulating quality measures for the healthcare sector.

     

Modeling undecided voters to forecast elections: From bandwagon behavior and the spiral of silence perspective

Forecasting election results has been a highly attractive activity among political and social scientists. Different forecasting methods have been proposed, but those based on public opinion polls are the most common. However, there are challenges to using opinion polls, especially because they neglect undecided voters. Due to the significant number of undecided participants and their impact on voting outcomes, we analyze the potential behavior of undecided voters by considering opinion polls and sentiment based on voter expectation from the perspective of the bandwagon effect and the spiral of silence. We establish a hierarchical Bayesian forecasting model to predict voting results, and apply it to the 2016 United States presidential election and the 2016 Brexit referendum. The results of our model suggest that voting outcomes are more predictable when fully utilizing the impact of undecided voters. The results indicate that integrating aggregated polls into the hierarchical Bayesian framework is a strong predictor for forecasting outcomes, and they provide evidence for the influence of sentiment based on voter expectation in forecasting election results.     

Are day traders bias free?—evidence from internet stock message boards

This study addresses the issue whether day traders’ recommendations on stocks are biasfree. We test whether on average day traders’ “Hold” sentiment is skewed and different from a neutral opinion. Posted messages and mature text classifier technology provide a novel approach to analyze the content of these “Hold” sentiment postings among day traders. Findings indicate that the self-disclosed “Hold” sentiment conveys an optimistic opinion and significantly differs from neutral. These results help both investors and researchers to better understand day traders’ psychology and behaviors when they recommend stocks. The paper also provides insight into the construction of future online sentiment indexes based on stock message boards.     

Mining semantic features in current reports for financial distress prediction: Empirical evidence from unlisted public firms in China

It is difficult to predict the financial distress of unlisted public firms due to their longer disclosure cycle of accounting information and more inadequate continuity of market trading information compared to listed firms. In this paper, we propose a framework to predict the financial distress of unlisted public firms using current reports. Specifically, to better represent the meaning of current report texts, we propose a semantic feature extraction method based on a word embedding technology. Empirical results show that current reports contain more effective information for predicting the financial distress of unlisted public firms compared with periodic reports. In addition, semantic features extracted using our proposed method significantly improve the predictive performance, and their enhancing effect is superior to that of topic features and sentiment features. Our study also provides implications for stakeholders such as investors and creditors.     

The COVID-19 pandemic uncertainty,investor sentiment,and global equity markets: Evidence from the time-frequency co-movements

We use daily data of the Google search engine volume index (GSVI) to capture the pandemic uncertainty and examine its effect on stock market activity (return, volatility, and illiquidity) of major world economies while controlling the effect of the Financial and Economic Attitudes Revealed by Search (FEARS) sentiment index. We use a time–frequency based wavelet approach comprising wavelet coherence and phase difference for our empirical assessment. During the early spread of the COVID-19, our results suggest that pandemic uncertainty, and FEARS sentiment strongly co-move, and increased pandemic uncertainty leads to pessimistic investor sentiment. Furthermore, our partial wavelet analysis results indicate a synchronization relationship between pandemic uncertainty and stock market activities across G7 countries and the world market. Our results are robust to the inclusion of alternative pandemic fear measure in the form of equity market volatility infectious disease tracker. The pandemic uncertainty and associated sentiment implications could be one plausible reason for increased volatility and illiquidity in the market, and hence, policymakers should look upon this issue for the financial market stability perspective.     

Daily news sentiment and monthly surveys: A mixed-frequency dynamic factor model for nowcasting consumer confidence

Policymakers, firms, and investors closely monitor traditional survey-based consumer confidence indicators and treat them as an important piece of economic information. To obtain a daily nowcast of monthly consumer confidence, we introduce a latent factor model for the vector of monthly survey-based consumer confidence and daily sentiment embedded in economic media news articles. The proposed mixed-frequency dynamic factor model uses a Toeplitz correlation matrix to account for the serial correlation in the high-frequency sentiment measurement errors. We find significant accuracy gains in nowcasting survey-based Belgian consumer confidence with economic media news sentiment.     

Text-based crude oil price forecasting: A deep learning approach

This study proposes a new, novel crude oil price forecasting method based on online media text mining, with the aim of capturing the more immediate market antecedents of price fluctuations. Specifically, this is an early attempt to apply deep learning techniques to crude oil forecasting, and to extract hidden patterns within online news media using a convolutional neural network (CNN). While the news-text sentiment features and the features extracted by the CNN model reveal significant relationships with the price change, they need to be grouped according to their topics in the price forecasting in order to obtain a greater forecasting accuracy. This study further proposes a feature grouping method based on the Latent Dirichlet Allocation (LDA) topic model for distinguishing effects from various online news topics. Optimized input variable combination is constructed using lag order selection and feature selection methods. Our empirical results suggest that the proposed topic-sentiment synthesis forecasting models perform better than the older benchmark models. In addition, text features and financial features are shown to be complementary in producing more accurate crude oil price forecasts.     

Assessing and predicting small industrial enterprises’ credit ratings: A fuzzy decision-making approach

Corporate credit-rating assessment plays a crucial role in helping financial institutions make their lending decisions and in reducing the financial constraints of small enterprises. This paper presents a new approach for small industrial enterprises’ credit-rating assessment using fuzzy decision-making methods and then tests this novel approach using real bank loan data from 1820 small industrial enterprises in China. The procedure of the proposed rating approach includes (1) using triangular fuzzy numbers to quantify the qualitative evaluation indicators; (2) adopting a correlation analysis, univariate analysis, and stepping backward feature selection method to select the input features; (3) employing the best-worst method (BWM) combined with the entropy weight method (EWM), the fuzzy c-means algorithm and the technique for order of preference by similarity to ideal solution (TOPSIS) to classify small enterprises into different rating classes; and (4) applying the lattice degree of nearness to predict a new loan applicant’s rating. We also conduct 10-fold cross-validation to evaluate the predictive performance of our proposed approach. The predictive results demonstrate that our proposed data-processing and feature selection approaches have better accuracy than the alternative approaches in predicting default, offering bankers a new valuable rating system to assist their decision making.     

A web-based system architecture for ontology-based data integration in the domain of IT benchmarking

In the domain of IT benchmarking (ITBM), a variety of data and information are collected. Although these data serve as the basis for business analyses, no unified semantic representation of such data yet exists. Consequently, data analysis across different distributed data sets and different benchmarks is almost impossible. This paper presents a system architecture and prototypical implementation for an integrated data management of distributed databases based on a domain-specific ontology. To preserve the semantic meaning of the data, the ITBM ontology is linked to data sources and functions as the central concept for database access. Thus, additional databases can be integrated by linking them to this domain-specific ontology and are directly available for further business analyses. Moreover, the web-based system supports the process of mapping ontology concepts to external databases by introducing a semi-automatic mapping recommender and by visualizing possible mapping candidates. The system also provides a natural language interface to easily query linked databases. The expected result of this ontology-based approach of knowledge representation and data access is an increase in knowledge and data sharing in this domain, which will enhance existing business analysis methods.     

行业资产的Black-Litterman模型配置研究——基于社交网络情绪文本挖掘算法

研究目标:构建反映行业股价走势的基于社交网络文本挖掘算法的行业投资者情绪指标，并改善嵌入行业投资者情绪指标的Black-Litterman模型对资产的配置结果。研究方法:基于社交网络文本挖掘算法度量投资者情绪，运用主成分分析法构建行业投资者情绪指标，并嵌入Black-Litterman模型中构建投资者观点矩阵，确定行业资产配置比。研究发现:基于行业投资者情绪的BL模型有效提高了资产配置的日均收益率和夏普比率。实证结果在样本外验证(除受新冠疫情影响阶段)、暴涨暴跌阶段以及经过允许卖空和交易成本调整后仍稳健，进而证实了投资者情绪对资产组合有显著影响。研究创新:基于社交网络文本挖掘算法构建投资者情绪指数，解决了仅依赖于预期收益或历史数据的预测模型无法直观揭示投资者心理认知和行为的局限性问题，从一个崭新的视角科学地解决Black-Litterman模型中投资者观点的生成问题。研究价值:扩展了Black-Litterman模型理论体系研究，并推动了行为金融理论在资产配置中的应用。     

Predicting stock market crises using daily stock market valuation and investor sentiment indicators

The purpose of this paper is to develop a daily early warning system for stock market crises using daily stock market valuation and investor sentiment indicators. To achieve this goal, we use principal components analysis to propose a comprehensive index of daily market indicators that reflects stock market valuation and investor sentiment. Based on the comprehensive index, we employ a logit model with Ensemble Empirical Mode Decomposition to develop a daily early warning system for stock market crises. Finally, we apply the proposed system to the early warning for stock market crises in China. The in-sample forecasting results show that investor sentiment and the forecast horizon by Ensemble Empirical Mode Decomposition improve the forecasting performance of conventional early warning systems. The out-of-sample forecasting results indicate that the proposed warning system still has a good performance.     

Forecasting sales using online review and search engine data: A method based on PCA–DSFOA–BPNN

Traditional sales forecasting methods are mainly based on historical sales data, which result in a certain lag. The relationship between sales volume and its influencing factors is intricate and often non-linear. In view of this, we propose a novel product forecasting method using online reviews and search engine data. Firstly, a dictionary-based sentiment analysis method is developed to convert the textual review concerning each attribute of the product into the corresponding sentiment score. And by combining the prospect theory and relevant online review data, sentiment indices in each period are calculated. Subsequently, data of product-related Baidu search words with different lag orders are collected and screened by time difference correlation analysis. Finally, the forecast model, PCA–DSFOA–BPNN, is constructed by combining the principal component analysis (PCA), the back propagation neural network (BPNN), and the improved fruit fly optimization algorithm (DSFOA), in which sentiment indices, Baidu search data, and historical sales volume are input data. Taking the monthly sales forecast of 14 automobile models as a case study, we observe that the proposed forecast method can effectively improve the forecast accuracy with good robustness.     

How accurate are government forecasts of economic fundamentals? The case of Taiwan

A government’s ability to forecast key economic fundamentals accurately can affect business confidence, consumer sentiment, and foreign direct investment, among others. A government forecast based on an econometric model is replicable, whereas one that is not fully based on an econometric model is non-replicable. Governments typically provide non-replicable forecasts (or expert forecasts) of economic fundamentals, such as the inflation rate and real GDP growth rate.In this paper, we develop a methodology for evaluating non-replicable forecasts. We argue that in order to do so, one needs to retrieve from the non-replicable forecast its replicable component, and that it is the difference in accuracy between these two that matters. An empirical example to forecast economic fundamentals for Taiwan shows the relevance of the proposed methodological approach. Our main finding is that the undocumented knowledge of the Taiwanese government reduces forecast errors substantially.     

MORE THAN A FEELING: CONFIDENCE,UNCERTAINTY, AND MACROECONOMIC FLUCTUATIONS

Economists, observers, and policy-makers often emphasize the role of sentiment as a potential driver of the business cycle. In this paper, we provide three contributions to this debate. First, we give an overview of the recent literature on the nexus between sentiment (considering both confidence and uncertainty) and economic activity. Second, we review existing empirical measures of sentiment, in particular consumer confidence, stock market volatility (SMV) and Economic Policy Uncertainty (EPU), on monthly data for 27 countries, 1985–2016. Third, we identify some new stylized facts based on international evidence. While different measures are surprisingly lowly correlated on average in each country, they are typically highly positively correlated across countries, suggesting the existence of a global factor or sizeable international spillovers of sentiment. Consumer confidence has the closest co-movement with economic and financial variables, and most of the correlations are contemporaneous or forward-looking, consistent with the view that economic sentiment is indeed a driver of activity.     

Expert opinion versus expertise in forecasting

Expert opinion is an opinion given by an expert, and it can have significant value in forecasting key policy variables in economics and finance. Expert forecasts can either be expert opinions, or forecasts based on an econometric model. An expert forecast that is based on an econometric model is replicable, and can be defined as a replicable expert forecast (REF), whereas an expert opinion that is not based on an econometric model can be defined as a non-replicable expert forecast (Non-REF). Both REF and Non-REF may be made available by an expert regarding a policy variable of interest. In this paper, we develop a model to generate REF, and compare REF with Non-REF. A method is presented to compare REF and Non-REF using efficient estimation methods, and a direct test of expertise on expert opinion is given. The latter serves the purpose of investigating whether expert adjustment improves the model-based forecasts. Illustrations for forecasting pharmaceutical stock keeping unit (SKUs), where the econometric model is of (variations of) the autoregressive integrated moving average model (ARIMA) type, show the relevance of the new methodology proposed in the paper. In particular, experts possess significant expertise, and expert forecasts are significant in explaining actual sales.     

Diversification discount and investor sentiment

While previous research has linked the diversification discount to suboptimal managerial decisions, recent empirical work and methods have shown these relationships are not as strong. A rational learning framework indicates the diversification discount is related to economic activity. Building on this framework, we test and find support for the hypothesis that investor sentiment explains the diversification discount. Investor sentiment favors riskier firms when sentiment is high, thereby increasing returns and relative valuations. As a result, diversified firms imputed value based on these multiples leads to a larger diversification discount and reverses when sentiment falls.     

A systematic approach to prioritizing R&D projects based on customer-perceived value using opinion mining

As product development has recently emphasized user innovation, it should necessarily reflect customer-perceived value, as well as technological value itself. However, while previous studies for technology planning have focused on analyzing the potential of technology, they have not considered the customer-perceived value that technology can create in a new product. Therefore, this research suggests a new approach to assessing the level of technology and evaluating R&D projects, by investigating customer-perceived value on technology through the use of the structural equation model and opinion mining. For this, the assessment framework is developed in terms of technology, product quality, and customer satisfaction, respectively, by investigating a variety of databases on each factor. The relationship between technology level and customer satisfaction is analyzed, using structural equation model and opinion mining. Based on the results, a strategy for technology development is established through gap analysis and simulation, after selecting and evaluating technologies that need to be developed. The proposed approach is applied to the real case of a moving system, in particular an automobile door, and we obtained that an R&D project for hinge-related technology would be promising, enhancing the customer satisfaction. It can suggest a future direction for new technology development. This paper contributes to enhancing the efficiency of technology planning based on the customer-perceived value, enabling the launch of new R&D projects.     

Forecasting Bitcoin with technical analysis: A not-so-random forest?

This paper uses data sampled at hourly and daily frequencies to predict Bitcoin returns. We consider various advanced non-linear models based on a multitude of popular technical indicators that represent market trend, momentum, volume, and sentiment. We run a robust empirical exercise to observe the impact of forecast horizon, model type, time period, and the choice of inputs (predictors) on the forecast performance of the competing models. We find that Bitcoin prices are weakly efficient at the hourly frequency. In contrast, technical analysis combined with non-linear forecasting models becomes statistically significantly dominant relative to the random walk model on a daily horizon. Our comparative analysis identifies the random forest model as the most accurate at predicting Bitcoin. The estimated measures of the relative importance of predictors reveal that the nature of investing in the Bitcoin market evolved from trend-following to excessive momentum and sentiment in the most recent time period.     

Demographics, dividend clienteles and the dividend premium

The catering theory of dividends proposed that corporate dividend policy is driven by prevailing investor demand for dividend payers, and that managers cater to investors by paying dividends when the dividend premium is high. While earlier research found that the dividend premium is not driven by traditional clienteles derived from market imperfections such as taxes, transaction costs, or institutional investment constraints, we find empirical evidence that demographic clienteles are an important source of the time-varying demand for dividend payers. In particular, we find that, as consistent with the behavioural life-cycle theory and the marginal opinion theory of stock price, the dividend premium is positively driven by demographic clientele variation represented by changes in the proportion of the older population. Our results are robust when controlled for the factors of investor sentiment, signalling, agency costs, tax clienteles, time trend, business cycle fluctuations and varying sample periods.     

The Triple Bottom Line: A Critical Review from a Transdisciplinary Perspective

The triple bottom line (TBL) has reformed management discourse by making sustainability part of the business agenda, yet increasingly the TBL has evolved into a proxy for sustainability, often visually depicted as the mutual maximization of economic, social and environmental dimensions. We use a sentiment analysis to show that the extant literature views the TBL favorably and uncritically, with only 8% of academic studies invoking the term negatively. Next, based on extant management literature, we show that two core assumptions underpin the TBL metaphor: win–win and firm‐level sustainability. Then we employ a transdisciplinary comparative analysis to contrast these assumptions with two ecological perspectives: strong sustainability and nested hierarchy. By drawing extensively from the literature of ecologically grounded sciences, our study contributes a critical evaluation of the TBL paradigm of sustainability.