Search results
1 – 4 of 4Daniel Šandor and Marina Bagić Babac
Sarcasm is a linguistic expression that usually carries the opposite meaning of what is being said by words, thus making it difficult for machines to discover the actual meaning…
Abstract
Purpose
Sarcasm is a linguistic expression that usually carries the opposite meaning of what is being said by words, thus making it difficult for machines to discover the actual meaning. It is mainly distinguished by the inflection with which it is spoken, with an undercurrent of irony, and is largely dependent on context, which makes it a difficult task for computational analysis. Moreover, sarcasm expresses negative sentiments using positive words, allowing it to easily confuse sentiment analysis models. This paper aims to demonstrate the task of sarcasm detection using the approach of machine and deep learning.
Design/methodology/approach
For the purpose of sarcasm detection, machine and deep learning models were used on a data set consisting of 1.3 million social media comments, including both sarcastic and non-sarcastic comments. The data set was pre-processed using natural language processing methods, and additional features were extracted and analysed. Several machine learning models, including logistic regression, ridge regression, linear support vector and support vector machines, along with two deep learning models based on bidirectional long short-term memory and one bidirectional encoder representations from transformers (BERT)-based model, were implemented, evaluated and compared.
Findings
The performance of machine and deep learning models was compared in the task of sarcasm detection, and possible ways of improvement were discussed. Deep learning models showed more promise, performance-wise, for this type of task. Specifically, a state-of-the-art model in natural language processing, namely, BERT-based model, outperformed other machine and deep learning models.
Originality/value
This study compared the performance of the various machine and deep learning models in the task of sarcasm detection using the data set of 1.3 million comments from social media.
Details
Keywords
Elisabetta Colucci, Francesca Matrone, Francesca Noardo, Vanessa Assumma, Giulia Datola, Federica Appiotti, Marta Bottero, Filiberto Chiabrando, Patrizia Lombardi, Massimo Migliorini, Enrico Rinaldi, Antonia Spanò and Andrea Lingua
The study, within the Increasing Resilience of Cultural Heritage (ResCult) project, aims to support civil protection to prevent, lessen and mitigate disasters impacts on cultural…
Abstract
Purpose
The study, within the Increasing Resilience of Cultural Heritage (ResCult) project, aims to support civil protection to prevent, lessen and mitigate disasters impacts on cultural heritage using a unique standardised-3D geographical information system (GIS), including both heritage and risk and hazard information.
Design/methodology/approach
A top-down approach, starting from existing standards (an INSPIRE extension integrated with other parts from the standardised and shared structure), was completed with a bottom-up integration according to current requirements for disaster prevention procedures and risk analyses. The results were validated and tested in case studies (differentiated concerning the hazard and type of protected heritage) and refined during user forums.
Findings
Besides the ensuing reusable database structure, the filling with case studies data underlined the tough challenges and allowed proposing a sample of workflows and possible guidelines. The interfaces are provided to use the obtained knowledge base.
Originality/value
The increasing number of natural disasters could severely damage the cultural heritage, causing permanent damage to movable and immovable assets and tangible and intangible heritage. The study provides an original tool properly relating the (spatial) information regarding cultural heritage and the risk factors in a unique archive as a standard-based European tool to cope with these frequent losses, preventing risk.
Details
Keywords
Jacob Mhlanga, Theodore C. Haupt and Claudia Loggia
This paper aims to explore the intellectual structure shaping the circular economy (CE) discourse within the built environment in Africa.
Abstract
Purpose
This paper aims to explore the intellectual structure shaping the circular economy (CE) discourse within the built environment in Africa.
Design/methodology/approach
The study adopted a bibliometric analysis approach to explore the intellectual structure of CE in the built environment in Africa. The authors collected 31 papers published between 2005 and 2021 from the Scopus database and used VOSviewer for data analysis.
Findings
The findings show that there are six clusters shaping the intellectual structure: demolition, material recovery and reuse; waste as a resource; cellulose and agro-based materials; resilience and low-carbon footprint; recycling materials; and the fourth industrial revolution. The two most cited scholars had three publications each, while the top journal was Resources, Conservation and Recycling. The dominant concepts included CE, sustainability, alternative materials, waste management, lifecycle, demolition and climate change. The study concludes that there is low CE research output in Africa, which implies that the concept is either novel or facing resistance.
Research limitations/implications
The data were drawn from one database, Scopus; hence, adoption of alternative databases such as Web of Science, Google Scholar and Dimensions could potentially have yielded a higher number of articles for analysis which potentially would result in different conclusions on the subject understudy.
Originality/value
This study made a significant contribution by articulating the CE intellectual structure in the built environment, identified prominent scholars and academic platforms responsible for promoting circularity in Africa.
Details
Keywords
Krištof Kovačič, Jurij Gregorc and Božidar Šarler
This study aims to develop an experimentally validated three-dimensional numerical model for predicting different flow patterns produced with a gas dynamic virtual nozzle (GDVN).
Abstract
Purpose
This study aims to develop an experimentally validated three-dimensional numerical model for predicting different flow patterns produced with a gas dynamic virtual nozzle (GDVN).
Design/methodology/approach
The physical model is posed in the mixture formulation and copes with the unsteady, incompressible, isothermal, Newtonian, low turbulent two-phase flow. The computational fluid dynamics numerical solution is based on the half-space finite volume discretisation. The geo-reconstruct volume-of-fluid scheme tracks the interphase boundary between the gas and the liquid. To ensure numerical stability in the transition regime and adequately account for turbulent behaviour, the k-ω shear stress transport turbulence model is used. The model is validated by comparison with the experimental measurements on a vertical, downward-positioned GDVN configuration. Three different combinations of air and water volumetric flow rates have been solved numerically in the range of Reynolds numbers for airflow 1,009–2,596 and water 61–133, respectively, at Weber numbers 1.2–6.2.
Findings
The half-space symmetry allows the numerical reconstruction of the dripping, jetting and indication of the whipping mode. The kinetic energy transfer from the gas to the liquid is analysed, and locations with locally increased gas kinetic energy are observed. The calculated jet shapes reasonably well match the experimentally obtained high-speed camera videos.
Practical implications
The model is used for the virtual studies of new GDVN nozzle designs and optimisation of their operation.
Originality/value
To the best of the authors’ knowledge, the developed model numerically reconstructs all three GDVN flow regimes for the first time.
Details