Search

Corporations develop, own, operate, and direct a majority of all assets in the built environment (BE), and the development and operation of these assets is a significant component of global sustainability. In recent years corporate sustainability action for large companies has trended toward becoming embedded within corporate environmental, social, and governance (ESG) programs. However, ESG strategies and their impacts are only beginning to be understood by large corporations, with little definition for how to apply, measure, and assess the effectiveness of ESG practices. Every industry sector has some connection to the BE, and construction engineering, and management (CEM) professionals contribute to all phases of the BE life cycle, including development, design, construction, and operations. As such, CEM professionals are vital in developing and implementing sustainability practices specifically for corporations with large footprints in the BE. The purpose of this research was to assess the extent to which the BE is impacted by corporate sustainability practice through ESG programs. The research utilized a grounded theory approach to explore sustainability practices directly associated with the BE from companies within all 11 Global Industry Classification Standard (GICS) sectors. It presents a representative model that encapsulates current organizational ESG strategy, through the development of ESG action categories. A total of 24 different ESG action categories are presented and defined. The research further identifies those ESG actions commonly applied by large corporations and how these actions differ by GICS sector. This research indicates that over half of corporate ESG actions are related to the BE. Ultimately, this paper demonstrates the necessary role that CEM professionals should play in influencing sustainability practice through ESG strategy and action within the BE. Practical Applications: ESG is a rising topic in corporate sustainability and is becoming increasingly tied to financial instruments such as sustainable investment indexes or green bonds. The rise of ESG presents both a challenge and an opportunity for CEM professionals. Large entities are expected to publish ESG reports sharing their sustainability progress, including entities in which CEM professionals are key figures, such as power utilities. The advent of ESG reporting presents a challenge to CEM professionals because they need to learn to navigate ESG reporting or risk losing access to valuable funding opportunities when subject to ESG measurement and evaluation. It also presents an opportunity for CEM professionals because they are uniquely qualified to lead ESG efforts as BE experts. ESG and the BE are strongly related, with most ESG work being environmental in nature. CEM professionals have been actively involved in improving the sustainability of BE assets as they relate to the environment for many years. To leverage this experience, CEM professionals need to take an active part in recommending improvements to corporate ESG strategy. Corporations themselves must also actively seek to engage CEM professionals in their ESG decision making processes and advisory boards. [ABSTRACT FROM AUTHOR]

The electronic nose is a non-invasive technology suitable for the analysis of edible oils. One of the practical applications in the olive oil industry is the classification of virgin oils based on their sensory characteristics. Notwithstanding that this technology, at this stage, cannot realistically replace the currently used methods, it is fruitful for a preliminary analysis of the oil quality. This work makes use of this technology to develop a methodology for the detection of the threshold by which an extra-virgin olive oil (EVOO) drops into the virgin olive oil (VOO) category. With this aim, two features were studied: the level of fruitiness level and the type of defect. The results showed a greater influence of the level of fruitiness than the type of defect in the determination of the detection threshold. Furthermore, three of the sensors (S2, S7 and S9) of the commercial e-nose PEN3 were identified as the most discriminating in the classification between EVOO and VOO oils. [ABSTRACT FROM AUTHOR]

Business parks have increasingly become part of Cape Town's industrial built environment over the last 25 years. It signifies a shift from large industrial premises to smaller units in communal industrial settings. This paper investigates the types of business activities found in the business parks using the conceptual underpinnings of location theory and industrial cluster theory. Online and fieldwork surveys of the businesses found in 49 business parks in Cape Town, South Africa, were conducted to compile a database of businesses and their activities. Each business was classified using the South African Standard Industrial Classification (SIC) codes to establish which economic sectors are represented in the business parks. It was found that the predominant economic sectors in business parks were the wholesale and retail trade, which included the repair of motor vehicles, motorcycles and personal and household goods and catering and accommodation, as well as the financial sector consisting of banking, insurance, real estate and business services. Manufacturing activities and commercial, social and personal services were also well represented. The findings indicate that business parks are the popular location for retail-orientated businesses and small manufacturing concerns and affirm the growing importance of the professional and business service sectors. The paper lays the groundwork for further investigation into the multi-scalar economic linkages of firms in business parks in Cape Town. [ABSTRACT FROM AUTHOR]

Social scientists, geographers, criminologists, and health scientists are often tasked with finding data to best capture the impact of "community context" on individual outcomes, including residential services, physical resources, and social institutions. One outlet for such data in Canada is Digital Map Technologies Inc. (DMTI) Spatial, which offers a national repository of over one million businesses and recreational points of interest. The database is generated through CanMap Streetfiles, which includes geocodes of each point's precise location. These data are available to researchers from their university data library and Esri Canada, but primarily available to private sector and government markets. That said, the goal of the current paper is to encourage researchers to access this rich yet under‐utilized data source. Each service, business, or resource in the DMTI Spatial database is assigned to a respective category using Standard Industrial Classification codes and North American Industrial Classification System codes. It is not clear, however, which is the more reliable coding criteria. We provide an overview of our review of DMTI Spatial data and take‐away suggestions for using this valuable resource for future research on meso‐level residential markers. Key messages: The goal of this paper is to outline existing data source(s) and measures from DMTI Spatial that might help capture meso‐level residential institutions.We recommend "best practices" for using DMTI Spatial data in researchers' own work to capture neighbourhood resources/amenities, or the social infrastructure of the community using either Standard Industrial Classification codes or North American Industrial Classification System codes.We conclude that Standard Industrial Classification codes in DMTI Spatial enhanced points of interest data are more complete—and more accurate—than North American Industrial Classification System codes. [ABSTRACT FROM AUTHOR]

This paper presents a broad study on the application of the BERT (Bidirectional Encoder Representations from Transformers) model for multiclass text classification, specifically focusing on categorizing business descriptions into 1 of 13 distinct industry categories. The study involved a detailed fine-tuning phase resulting in a consistent decrease in training loss, indicative of the model's learning efficacy. Subsequent validation on a separate dataset revealed the model's robust performance, with classification accuracies ranging from 83.5% to 92.6% across different industry classes. Our model showed a high overall accuracy of 88.23%, coupled with a robust F1 score of 0.88. These results highlight the model's ability to capture and utilize the nuanced features of text data pertinent to various industries. The model has the capability to harness real-time web data, thereby enabling the utilization of the latest and most up-to-date information affecting to the company's product portfolio. Based on the model's performance and its characteristics, we believe that the process of relative valuation can be drastically improved. [ABSTRACT FROM AUTHOR]

The task of company classification is traditionally performed using established standards, such as the Global Industry Classification Standard (GICS). However, these approaches heavily rely on laborious manual efforts by domain experts, resulting in slow, costly, and vendor-specific assignments. Therefore, we investigate recent natural language processing (NLP) advancements to automate the company classification process. In particular, we employ and evaluate various NLP-based models, including zero-shot learning, One-vs-Rest classification, multi-class classifiers, and ChatGPT-aided classification. We conduct a comprehensive comparison among these models to assess their effectiveness in the company classification task. The evaluation uses the Wharton Research Data Services (WRDS) dataset, consisting of textual descriptions of publicly traded companies. Our findings reveal that the RoBERTa and One-vs-Rest classifiers surpass the other methods, achieving F1 scores of 0.81 and 0.80 on the WRDS dataset, respectively. These results demonstrate that deep learning algorithms offer the potential to automate, standardize, and continuously update classification systems in an efficient and cost-effective way. In addition, we introduce several improvements to the multi-class classification techniques: (1) in the zero-shot methodology, we TF-IDF to enhance sector representation, yielding improved accuracy in comparison to standard zero-shot classifiers; (2) next, we use ChatGPT for dataset generation, revealing potential in scenarios where datasets of company descriptions are lacking; and (3) we also employ K-Fold to reduce noise in the WRDS dataset, followed by conducting experiments to assess the impact of noise reduction on the company classification results. [ABSTRACT FROM AUTHOR]

Digital technology innovation is the core driving force for the high-quality development of the digital economy, and in-depth exploration of the regional distribution pattern and formation mechanism of digital technology innovation in China is conducive to the rational layout and coordinated development of the inter-provincial digital economy. Based on the Reference Relationship Table of the Classification of Core Industries of Digital Economy and the International Patent Classification (2023), the patent authorization data of digital technology from 2012 to 2022 were obtained, and the spatiotemporal situation of China's digital technology innovation was analyzed by using ArcGIS software, Dagum's Gini coefficient, and Moran's I index, and the spatial Dubin panel model was used to explore the influencing factors of digital technology innovation. It is found that: (1) the scale and vitality of China's digital technology innovation have increased significantly, and there are obvious spatial differentiation characteristics, and the innovation level of "eastern coastal—central and western interior" is decreasing, forming a cluster distribution pattern in the Yangtze River Delta region, Beijing, Guangdong, and other places, and the degree of agglomeration is decreasing. (2) The overall regional differences in China's digital technology innovation are large, the differences between the East and the West dominate the interregional differences, and the net differences between regions are the main factors leading to regional differences. (3) There is a significant positive spatial correlation between the scale and vitality of digital technology innovation, which has a significant spatial spillover effect. (4) The results confirm that the level of economic development, digital access, financial scientific and technological support, technology market development level, and R&D intensity have a significant positive impact on the scale and vitality of digital technology innovation; The investment in scientific and technological talents has a significant positive impact on the scale of digital technology innovation, but has no significant impact on the vitality of digital technology innovation. [ABSTRACT FROM AUTHOR]

Endometriosis is a chronic and debilitating condition that affects daily working life. Characterization of the factors associated with endometriosis in the working population can facilitate the development of prevention and intervention strategies for those at risk of endometriosis. This population-based retrospective study was conducted using the 2007–2015 National Health Insurance Service–Female Employees database. Overall, 151,386 female workers aged 15–64 years were included in the study. Participants with endometriosis were identified using the diagnosis codes in the claims data. Multivariable Cox regression analyses were used to evaluate the effect of sociodemographic, lifestyle, health, and occupational factors on endometriosis risk. Of the 151,386 participants, 4,457 were diagnosed with endometriosis. The risk of endometriosis was significantly higher in 41–60 years group (HR = 1.47 (95% CI, 1.06–2.04)) and in those with body mass index (BMI) < 18.5 kg/m2 (HR = 1.16 (95% CI, 1.05–1.27)) than 15–20 years group and those with normal BMI, respectively. According to the international standard industrial classification, occupational groups with financial and insurance activities, public administration and defence, compulsory social security, and manufacturing were at a higher risk of endometriosis. Although there was no significant association between the risk of endometriosis and type of work, the cumulative prevalence of endometriosis from 2007 to 2015 continued to rise in office workers, manual workers, and both types of workers together. The risk of endometriosis was closely linked to the occupational characteristics of female workers. This study provides a foundation for developing occupational safety and health guidelines for female workers. [ABSTRACT FROM AUTHOR]

The article investigates how declining immigration affects online job postings in the U.S., particularly in labor markets reliant on immigrant workers. As immigration decreased from 2016 to 2021 amid labor shortages, the study finds that in immigrant-reliant markets, job search efforts increased, reflected in online postings with higher starting wages and slower-growing skill requirements.

Interest has focused on quantifying the size and scope of environmental markets, particularly those that offset ecosystem impacts or restore natural infrastructure to improve habitat or promote clean air and water. In this paper, we focus on the US wetland and stream compensatory mitigation market, asking: what types of firms make up the mitigation "industry"? What are the economic impacts–i.e., the "size"–of the mitigation industry? How has this industry changed over time? We present the results of a national survey of mitigation firms and construct an input-output model of the industry's economic impacts and employment. We also develop a comparative, 2014 model of the industry using data from a previous study of the broader, ecological restoration economy. Our findings suggest that the (2019, pre-COVID) mitigation industry collects annual revenues (direct economic impacts) in excess of $3.5 billion, which, along with additional indirect (supply chain) and induced (spillover) economic impacts, combine to over $9.6 billion in total output and support over 53,000 total jobs. We estimate 2014–2019 growth of ~35.2 percent in revenues, ~32.6 percent in total economic impacts, and a compound annual growth rate (CAGR) of 5.25%. This places the mitigation industry within the range of other, well-established industries within the technical services sector. We suggest establishing North American Industry Classification System (NAICS) codes specifically for ecological restoration and mitigation firms, an essential step in generating accurate and consistent employment estimates in the future, particularly at sub-national geographic scales. [ABSTRACT FROM AUTHOR]

Industry assignment, which assigns firms to industries according to a predefined industry classification system (ICS), is fundamental to a large number of critical business practices, ranging from operations and strategic decision-making by firms to economic analyses by government agencies. Three types of expert knowledge are essential to effective industry assignment: definition-based knowledge (i.e., expert definitions of each industry), structure-based knowledge (i.e., structural relationships among industries as specified in an ICS), and assignment-based knowledge (i.e., prior firm-industry assignments performed by domain experts). Existing industry assignment methods utilize only assignment-based knowledge to learn a model that classifies unassigned firms to industries, overlooking definition-based and structurebased knowledge. Moreover, these methods only consider which industry a firm has been assigned to, ignoring the time-specificity of assignment-based knowledge, i.e., when the assignment occurs. To address the limitations of existing methods, we propose a novel deep learning-based method that not only seamlessly integrates the three types of knowledge for industry assignment but also takes the timespecificity of assignment-based knowledge into account. Methodologically, our method features two innovations: dynamic industry representation and hierarchical assignment. The former represents an industry as a sequence of time-specific vectors by integrating the three types of knowledge through our proposed temporal and spatial aggregation mechanisms. The latter takes industry and firm representations as inputs, computes the probability of assigning a firm to different industries, and assigns the firm to the industry with the highest probability. We conduct extensive evaluations with two widely used ICSs and demonstrate the superiority of our method over prevalent existing methods. [ABSTRACT FROM AUTHOR]

Conventional wisdom holds that North American Industry Classification System (NAICS) codes chosen by people not experienced with the system are often mis-specified, but there has been little formal research into the scope of the problem. In this paper we explore prevalence of and patterns in misspecification in NAICS codes self-reported on two kinds of business tax forms. Errors are identified by comparing as-filed codes with codes validated by Statistics of Income. We find that over a third of codes are wrong, but that the errors are not random and often (though not always) seem to have logical reasons behind them. [ABSTRACT FROM AUTHOR]

The article focuses on the Institute for Supply Management's quarterly forecast, which predicts spending increases for printing and paper manufacturing. It mentions the forecast indicates that manufacturing and services sectors are expected to experience revenue growth, while capital expenditures and prices paid for raw materials are projected to increase.

Recently, various sophisticated methods, including machine learning and artificial intelligence, have been employed to examine health-related data. Medical professionals are acquiring enhanced diagnostic and treatment abilities by utilizing machine learning applications in the healthcare domain. Medical data have been used by many researchers to detect diseases and identify patterns. In the current literature, there are very few studies that address machine learning algorithms to improve healthcare data accuracy and efficiency. We examined the effectiveness of machine learning algorithms in improving time series healthcare metrics for heart rate data transmission (accuracy and efficiency). In this paper, we reviewed several machine learning algorithms in healthcare applications. After a comprehensive overview and investigation of supervised and unsupervised machine learning algorithms, we also demonstrated time series tasks based on past values (along with reviewing their feasibility for both small and large datasets). [ABSTRACT FROM AUTHOR]

The major sectors exainined include: Agriculture, Construction and Mining Manufacturing Retail Accommodation and Food Services Transport and Logistics Art and Communications Financial and Insurance Services Local Government Higher Education Healthcare Utilities Specific questions/areas explored include a SWOT analysis, the impact of technologies on the sector, and each sector's preparedness, including policies and practices, not least skill requirements for the emerging workplace. A point I concur with - as my colleague Chris Brewster and I recently wrote on this topic regarding work and employment - is that jobs are being replaced by artificial intelligence (AI) in this period, but this AI is being replaced by more advanced AI. The key conclusions from this research are a call for a national technology policy, investment in AI development, including training, education and development. [Extracted from the article]

However, the percentage of board members who are clinicians has recently declined.[4] In 2018, almost one-third of hospital boards did not have a single physician member, while almost two-thirds lacked a single nurse. DISCUSSION At top-ranked US hospitals, the most common professional background for board members is finance, far exceeding representation from physicians, nurses, and other health care workers. First, we studied a limited sample of top-ranked hospitals which may not be representative of all hospitals. [Extracted from the article]

This paper analyzes the differences in productivity performance between exporting and non-exporting manufacturing firms in Ecuador during the period 2007–2018, using two underexplored industry classifications in this analysis: the Pavitt Taxonomy and the OECD technological intensity classification. We estimate total factor productivity (TFP) at the firm level with a modified (Gandhi, Navarro, and Rivers 2020) approach that allows us to use an additional state variable and endogenize the law of motion for productivity, which allows the past decision about exporting to affect future productivity. This approach captures the static and dynamic gains (learning-by-exporting) in productivity from exporting. Moreover, we test self-selection on the entry and exit sides of the market, and show a robustness check of the learning-by-exporting hypothesis by using a difference-in-difference matching estimator. Our results indicate that exporters have better productivity performance than non-exporters. We find evidence in favor of the self-selection on the entry and exit sides of the market. Finally, we show there are dynamic gains in productivity from exporting, supporting the learning-by-exporting hypothesis. [ABSTRACT FROM AUTHOR]

Purpose: This paper aims to evaluate whether the integration of smart specialization strategies (S3) into clusters significantly impacts their efficiency for countries that still do not implement this policy. This study tests three effects: whether the kind of policies envisaged through an S3 strategy impacts cluster's efficiency; whether this impact changes with the technological intensity of the clusters; to determine which S3 is more suitable for sub-clusters at different levels of technological intensity. Design/methodology/approach: The Mexican economy is taken as case of study because it has a proper classification of its industries intro Porter's cluster's definition but still does not adopt the S3 policy. Through data envelopment analysis (DEA), this study evaluates the cluster's efficiency increment when variables representing the S3 elements are included. Findings: The results show that strategies following the S3 had a significant impact in all clusters, but when clusters were classified by technological intensity, the impact on efficiency is higher in clusters in the medium low-tech group. Practical implications: According to the results in the DEA, it can be concluded that these S3 strategies have the potential to increase the clusters' productivity significantly. These results make convenient the adoption of the S3 policy by countries that already count with a properly cluster definition. Originality/value: These findings contribute to the lack of studies that analyze the join implementation of S3 on clusters. [ABSTRACT FROM AUTHOR]

Background: The important role of mental health in sustainable economic development is gradually being recognized. This study aimed to evaluate the prevalence and gender disparity of those who screen positive for depression in China by the employer and industrial classification. Methods: We used data from a nationally representative survey, the China Family Panel Studies. Depression was judged by the Centre for Epidemiologic Studies Depression Scale. Employer classifications were categorized according to the local characteristics of Mainland China. Industrial classifications were defined using level-1 of the China version of the International Standard Industrial Classification of All Economic Activities. Weighted logistic regressions were fitted to estimate the gender disparities, controlling for confounders. Results: Forty eight thousand six hundred twenty eight adults were included. 18.7% (95%CI 18.1–19.4) of sampled adults were screened positive for depression symptoms, with 16.6% (95%CI 15.8–17.5) in males vs 21.0% (95%CI 20.1–22.0) in females. By classification of the employer, the prevalence was lowest among those employed by Government/party organisations (11.8%, 95%CI 8.9–15.4), and highest in those self-employed (21.8%, 95%CI 20.8–22.9); the gender disparity was mainly found in those employed by Sole proprietorship (Adjusted odds ratio [AOR] = 1.95, 95%CI 1.19–3.19) and Private enterprise (AOR = 1.34, 95%CI 1.13–1.59), as well as those self-employed (AOR = 1.49, 95%CI 1.3–1.17). By industrial classification, the prevalence was lowest among those who worked in the industry of Real estate (7.2%, 95%CI 4.8–10.6), and highest among those who worked in the industry of Agriculture, forestry, animal husbandry and fishing (22.9%, 95%CI 15.5–32.4); the gender disparity was mainly found in those who worked in the industry of Agriculture, forestry, animal husbandry and fishing (AOR = 3.29, 95%CI 1.18–9.15), Manufacturing (AOR = 1.41, 95% CI 1.09–1.82), Wholesale and retail trade (AOR = 1.48, 95% CI 1.07–2.06), and Accommodation and food service (AOR = 1.91, 95% CI 1.15–3.18). Conclusion: The prevalence of depression in China had a wide variation by classifications of the employer and industry. Gender disparities were identified among workers from Sole proprietorship, Private enterprise, and self-employed, or workers from the industry of Agriculture, forestry, animal husbandry and fishing, Manufacturing, Wholesale and retail trade, and Accommodation and food service. [ABSTRACT FROM AUTHOR]

In particular, three main topics are analyzed: (i) FDD classification and taxonomy, (ii) approaches to data-driven FDD in HVAC systems, (iii) deployment of FDD strategies in buildings and related impact assessment. The main goal of Fault Detection and Diagnosis (FDD) processes is to identify faults, determine their sources, and recognize solutions before the system is further harmed or service is lost. In the case where model-based methodologies are adopted, professionals can develop the FDD tools by using only the building or system's metadata, while data-based methods require calibrated measurement training data. Kim I. and Kim W. [[6]] presented a data-driven FDD approach that uses ML classification methods to detect and diagnose faults in a 90 ton (approximately 316 kW) centrifugal chiller system. [Extracted from the article]

Objective: to analyze the challenges faced by nurses in the risk classification of an urgency and emergency service. Method: exploratory and descriptive research, with a qualitative approach. Data collection took place in March 2019, through semistructured and individual interviews with nurses working in an Emergency Care Unit in the city of Caruaru, Pernambuco, Brazil. For the analysis, Bardin's content analysis was used. Results: three categories were generated: nursing care in risk classification; challenges of the risk classification industry, and challenges of the tool that defines the risk classification. Participants pointed out issues such as lack of understanding of the population, outdated protocol and disagreement with other team members, as the main challenges. Conclusion: it was found that the role of nursing in risk classification is still a complex process, which does not strictly depend on protocols, but on motivated professionals, continuously trained and in tune. [ABSTRACT FROM AUTHOR]

We conducted the first profitability comparison study across health care industries in the United States, using the DuPont Analysis framework. The combination of Return on Equity (ROE) and ROE volatility was used to provide a comprehensive "risk-return" approach for profitability comparison. Based on the 2010–2019 financial disclosures of 1,231 publicly traded health care companies in the U.S. that reported positive assets and equity, we estimated the industry-specific fixed effects on ROE and its three components—profit margin, asset utilization, and financial leverage—for ten industries in the health care sector, classified by the Global Industry Classification Standard (GICS). For each industry, we also estimated its fixed effects on ROE volatility. We found that the pharmaceuticals industry and biotechnology industry have lower ROE—mainly driven by their relatively low profit margin and low assets utilization—and higher ROE volatility than other health care industries. We also found that the health care facilities industry relies most on debt financing. This study demonstrates a holistic approach for profitability comparison across industries. [ABSTRACT FROM AUTHOR]

The aroma of wine is traditionally analyzed by sensory methods or by using gas chromatography; both analytical methodologies are slow and expensive and do not allow continuous monitoring. For this reason, interest in rapid methods has increased in recent times. Electronic noses (e-noses) stand out for their high sensitivity, speed, low cost, and little or no sample preparation. They present, however, low selectivity, which requires advance analytical methods to distinguish compounds. Here, we present a low-cost e-nose device for the analysis and identification of distinct varieties of wine. Chemical analysis data are compared to e-nose data through a principal component analysis (PCA) and a k-means clustering algorithm to establish relationships between varieties of wines and the e-nose classification capability. The results show that e-nose technology found significant differences between the analyzed samples, and furthermore, classifying the samples in accordance with the chemical analysis classification. The maximal accuracy obtained was 100% using the k-means algorithm for binary classification with N = 21 samples. Thus the potential of e-nose technology was shown in the wine industry for the identification and classification of wine varieties or quality. [ABSTRACT FROM AUTHOR]

Before purchasing a product or service, People typically analyse the information about the product such as cost, warranty, quality etc. Only after getting satisfaction about such things, People try to buy that product based on the quality of service received. Since this process takes time and a chance of being duped by the dealer are higher, Sentiment analysis (SA) is necessary to purchase a product without any hesitation. Sentiment analysis examines reviews and comments of the products, which are in the form of text that requires several processes for providing the desirable information to the People. Moreover, SA is a significant research direction of Natural Language Processing (NLP). In this paper, a novel sentiment analysis model is developed based on the Machine Learning (ML) Algorithm, which provides an accurate sentiment information for the texts having different perspectives. The method of Stop words are used for data pre-processing. By using count vectorizer, the text data is converted into the form of vectors for extracting the desired features. Finally, the type of sentiment whether it is positive, negative or neutral is determined based on the ML classifier namely, Naive Bayes classifier. This model is developed using Django web framework that provides an accurate sentiment classification to the people or the industries who need the sentiment analysis. [ABSTRACT FROM AUTHOR]