Improving Tourism Analytics From Climate Data Using Knowledge Graphs

Wu, Jiantao; Pierse, Jarrett; Orlandi, Fabrizio; O’Sullivan, Declan; Dev, Soumyabrata

doi:10.1109/jstars.2023.3239831

Cited by 6 publications

(4 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Integrating scientific equations into the loss functions of ML models can be especially useful when the goal is to solve a known target equation, e.g., in the problem of solving PDEs [4,14,176]. Similarly, heuristics, rules, and structured relationships in ontologies and knowledge graphs can also be encoded and added to the loss function [19,42,165,166,167]. By penalizing deviations from these established knowledge bases, the training algorithm has a higher chance of learning physically consistent data patterns and thus achieving improved generalizability.…”

Section: Kgml Methods Additional Commentsmentioning

confidence: 99%

Knowledge-Guided Machine Learning

Karpatne¹,

Kannan²,

Kumar³

2022

View full text Add to dashboard Cite

This paper presents an overview of scientific modeling and discusses the complementary strengths and weaknesses of ML methods for scientific modeling in comparison to process-based models. It also provides an introduction to the current state of research in the emerging field of scientific knowledge-guided machine learning (KGML) that aims to use both scientific knowledge and data in ML frameworks to achieve better generalizability, scientific consistency, and explainability of results. We discuss different facets of KGML research in terms of the type of scientific knowledge used, the form of knowledge-ML integration explored, and the method for incorporating scientific knowledge in ML. We also discuss some of the common categories of use cases in environmental sciences where KGML methods are being developed, using illustrative examples in each category.

show abstract

Section: Kgml Methods Additional Commentsmentioning

confidence: 99%

Knowledge-Guided Machine Learning

Karpatne¹,

Kannan²,

Kumar³

2022

View full text Add to dashboard Cite

show abstract

“…One of the most notable applications is its use in feature selection for machine learning tasks. Our previous studies [66], [79], [80] have examined triplestorebased KGs in terms of their ability to enable data consumers to acquire timestamped features from multiple data sources using SPARQL queries. This considerably reduces the I/O cost and preprocessing efforts during the training datasets preparation.…”

Section: B Enhanced Time Series Data Analyticsmentioning

confidence: 99%

On the Use of Virtual Knowledge Graphs to Improve Environmental Sensor Data Accessibility

Wu,

Orlandi,

O'Sullivan

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

The rapid proliferation of Environmental Sensor Networks (ESNs) used for monitoring environmental systems such as meteorology and air quality, as well as advances in database technologies (e.g., SQL) has made significant progress in sensor data management. Notwithstanding the strength of these databases, they can inevitably lead to a data heterogeneity problem as a result of isolated databases with distinct data schema, which are expensive to be accessed and pre-processed when the data is consumed spanning multiple databases. Recently, knowledge graphs have been used as one of the most popular integration frameworks to address this data heterogeneity problem from the perspective of establishing an interoperable semantic schema (a.k.a. ontology). However, the majority of proposed knowledge graphs in this domain are a product of an ETL (Extraction-Transform-Load) approach with all the data physically stored in a triplestore. In contrast, this paper examines an approach of virtualizing knowledge graphs on top of the SQL databases as the means to provide a federated data integration approach for enhanced heterogeneous Environmental Sensor Networks' data access, bringing with it the promise of more cost-efficiency in terms of I/O, storage, etc. In addition, this work also considers some motivating application scenarios regarding the efficiency of time series data access. Based on a performance comparison between the proposed integration approach and some popular triplestores, the proposed approach has a significant edge over triplestores in multiple time series structuring and acquisition.

show abstract

“…Wu et al [18] proposed a new tourism analytic technique using a knowledge graph. The actual goal of the technique is to identify the climatic change of a particular place using climate data.…”

Section: Related Workmentioning

confidence: 99%

Using the Fuzzy Method and Multi-Criteria Decision Making to Analyze the Impact of Digital Economy on Urban Tourism

Wang

2024

Int J Comput Intell Syst

View full text Add to dashboard Cite

Urban tourism promotes the economic growth of a nation around the year through direct and indirect incomes. In recent years, the digital economy has impacted the growth of urban tourism through hassle-free money transactions and expenditures. This article, therefore, introduces a Multi-Criteria Fuzzy-based Decision-Making Method (MCFDMM) for validating the impact of the digital economy impact over tourism. The study introduces a new framework, DLFDSS-RRM, that uses deep learning and fuzzy decision support systems for residence right management, enhancing resource allocation, security, and resident satisfaction in urban residential communities. The criteria such as expenses, positive response, and repeated payments are validated by the tourists across their travel plan. These conditions satisfying the tourist’s expectations are estimated based on their reviews of economic conditions are validated. The validation is performed against the growth of the country from urban tourism. The fuzzy process validates the growth of the country between two successive financial quarters based on the above conditions. In the condition analysis, the fuzzy process identifies the least derivatives contributing to minimal economic growth. This is reversed using the hiking condition that occurs in any quarter and hinders economic growth. Therefore, the process is validated using the metrics growth rate, condition satisfaction, analysis rate, analysis time, and unrelated assessment. The comparative analysis across various models reveals growth rates ranging from 0.263 to 0.4055, condition satisfaction percentages from 53.747 to 74.351, and analysis rates from 0.275 to 0.4662.

show abstract

Improving Tourism Analytics From Climate Data Using Knowledge Graphs

Cited by 6 publications

References 34 publications

Knowledge-Guided Machine Learning

Knowledge-Guided Machine Learning

On the Use of Virtual Knowledge Graphs to Improve Environmental Sensor Data Accessibility

Using the Fuzzy Method and Multi-Criteria Decision Making to Analyze the Impact of Digital Economy on Urban Tourism

Contact Info

Product

Resources

About