**2.9 Review of supply chain sector**

Antonio et al. presented a literature survey of simulation techniques in supply chain risks [41]. The authors have highlighted the significance of BDI in supply chain systems. This analysis has concluded that the problem at hand is simplified without complexity in modeling. This study has complied with industry 4.0 standards. Ostrowski et al. have explored the potential of semantic web technologies by demonstrating a case study in the supply chain [42]. Authors have identified the system for supporting data from multiple sources. This study was carried out by semi-automated mapping using shared domain ontology. Awwad et al. provided a review on applications, advantages and issues of BDI technologies for the supply chain management [43]. The supply chain risk management was carried out using data analytics by making a proactive decision. Lia and Liu have illustrated a data-driven framework for supply chain management [44]. The various circumstances of the supply chain


#### **Table 9.**

*Communication sector review summary.*


**Table 10.**

*Supply chain sector review summary.*

are accommodated by enabling multiple working modes. Benabdellah et al. discussed the impact of big data on supply chain management [45].

The survey of various supply chain operation reference model presented their applications and challenges. **Table 10** presents a summary of highlights from the supply chain sector review.

### **2.10 Review of research domain**

A review by Li presented BDI technology applications for the analysis of Chinese and Russian dance components with modern features [46]. Arputhamary and Arockiam presented the prominence of BDI by identifying the open problems and the same is extended to proceed with future research in the big data environment [47]. Kadadi et al. have surveyed BDI methods and their interoperability [48]. Authors have also explored its usage in big data setup and the corresponding challenges. Ostrowski and Kim have presented a BDI strategy based on ontology [49]. BDI strategy was implemented in Apache Spark prototyping environment that generates ontology versions using rule-based translations. Sottovia et al. have described the Research Alps project pipeline. This project was funded by the EU Commission [50]. They have created an open dataset providing Alpine area research centre details. Portugal et al. have presented a high-level spatial–temporal architectural framework for massive data integration, analysis and provenance management [51]. This methodology was applied for BDI analysis. **Table 11** presents a summary of the highlights of the research review.

#### **2.11 Review of recent advancements in BDI**

Large scale implementation of BDI solutions is a very complex and difficult process than automating data transformation processes. To reduce the complexity, the organizations should implement the procedures for data discovery, semantic or

**59**

*Big Data Integration Solutions in Organizations: A Domain-Specific Analysis*

46 Big Data Technology Combination of Chinese

**Ref. No. Methods Used Results Applications**

and Russian cultural and modern features

The existing techniques and approaches are inefficient to handle the

Big data integration architecture

Multiple data sources 'Semi-automated mapping

Open dataset providing Alpine area research centres details

Model-driven techniques resulting in data integration and

analysis

problems.

Dance elements

Possible research directions

BDI within the organization and inter organizations

Large scale Big data applications

Research Alps project funded by EU commission

Provenance information

business comprehension of data, metadata management, structured and unstructured data management, and transformation. Integrating unstructured and semistructured data enables organizations to manage modern data sources containing text, images, and video. A survey was conducted by AtScale Inc. in collaboration with Cloudera and ODPi.org reveals that most of the organizations are selecting multi-cloud strategies for BDI implementation. Data virtualization and data governance are their top priorities [52]. This survey has collected data from 150 data practitioners where the respondents are from multiple industries around the world. The online magazine "Smarter with Gartner" has reported that top ten technology

This article revealed that the combination of machine learning algorithms and data technologies could help the medical and public health experts to discover new possible treatments. The article entitled "2020 CRN Big Data 100" published in "Data Integration Solutions Review" enlists the emerging big data tool vendors [54]. This list provides the details of data integration software, tools, platforms and vendors. A data-driven technique for a hybrid BDI using multilayer perceptron was discussed in this research [55]. A customized multilayer perceptron model was constructed using time-based parameters. The fields applied in optimization analysis are also used in the error matrix through additional neural network model. Research results revealed that this solution captures the variations in state variables. BDI project implementation for COVID-19 analytics was discussed [56]. This project was funded by the European Union research fund. This platform combines information from multiple sources such as world news, social media, published science and health data from healthcare institutions. The project design was co-created with industry, academia, health professionals, and policymakers to align with innovative technologies. This project successfully

trends in data analytics require essential investments [53].

provides useful and actionable information to public health authorities.

*DOI: http://dx.doi.org/10.5772/intechopen.95800*

47 Importance of BDI issues and challenges are identified

48 Addressing challenges of BDI such as Data accommodation, Data irregularity, Query optimization, Extensibility, ETL processing

49 Ontology-based data integration,

50 M-STEP and entity matching

instances

*Research domain review summary.*

**Table 11.**

51 Domain experts focusing on

creation of new ontology versions by using rule-based translation

method and functional framework to deal with hierarchical data

appropriate analysis steps, high-level models linked with code

produces middleware


*Big Data Integration Solutions in Organizations: A Domain-Specific Analysis DOI: http://dx.doi.org/10.5772/intechopen.95800*

#### **Table 11.**

*Data Integrity and Quality*

41 Literature survey of

43 Detailed review on

management

44 Design and development of

45 A detailed survey of various supply chain operations reference model with opportunities and challenges.

42 Semantics with annotation in

simulation techniques for the analysis and synthesizing risks in supply chains

Ontology federation process, Integration of data from multiple resources using shared domain ontology

applications, advantages and issues of big data technologies for supply chain

a data-driven framework for supply chain management

are accommodated by enabling multiple working modes. Benabdellah et al. discussed

**Ref. No. Methods Used Results Applications**

supply chains

Analyzed the impact of risks in

Integration of data from multiple resources using semiautomated mapping

Infrastructure and human skillset need to be improved, new and effective techniques need to be developed

Multiple working modes of big data in supply chain

Studies revealed that the supply chain process is having

higher importance

management

Supply chain management

Supply chain risk detection

Supply chain in manufacturing and

Power split device in hybrid vehicles

The supply chain and manufacturing

products

logistics

The survey of various supply chain operation reference model presented their applications and challenges. **Table 10** presents a summary of highlights from the

A review by Li presented BDI technology applications for the analysis of Chinese

Arockiam presented the prominence of BDI by identifying the open problems and the same is extended to proceed with future research in the big data environment [47]. Kadadi et al. have surveyed BDI methods and their interoperability [48]. Authors have also explored its usage in big data setup and the corresponding challenges. Ostrowski and Kim have presented a BDI strategy based on ontology [49]. BDI strategy was implemented in Apache Spark prototyping environment that generates ontology versions using rule-based translations. Sottovia et al. have described the Research Alps project pipeline. This project was funded by the EU Commission [50]. They have created an open dataset providing Alpine area research centre details. Portugal et al. have presented a high-level spatial–temporal architectural framework for massive data integration, analysis and provenance management [51]. This methodology was applied for BDI analysis. **Table 11** presents a summary of the highlights of the research review.

Large scale implementation of BDI solutions is a very complex and difficult process than automating data transformation processes. To reduce the complexity, the organizations should implement the procedures for data discovery, semantic or

and Russian dance components with modern features [46]. Arputhamary and

the impact of big data on supply chain management [45].

supply chain sector review.

*Supply chain sector review summary.*

**Table 10.**

**2.10 Review of research domain**

**2.11 Review of recent advancements in BDI**

**58**

*Research domain review summary.*

business comprehension of data, metadata management, structured and unstructured data management, and transformation. Integrating unstructured and semistructured data enables organizations to manage modern data sources containing text, images, and video. A survey was conducted by AtScale Inc. in collaboration with Cloudera and ODPi.org reveals that most of the organizations are selecting multi-cloud strategies for BDI implementation. Data virtualization and data governance are their top priorities [52]. This survey has collected data from 150 data practitioners where the respondents are from multiple industries around the world. The online magazine "Smarter with Gartner" has reported that top ten technology trends in data analytics require essential investments [53].

This article revealed that the combination of machine learning algorithms and data technologies could help the medical and public health experts to discover new possible treatments. The article entitled "2020 CRN Big Data 100" published in "Data Integration Solutions Review" enlists the emerging big data tool vendors [54]. This list provides the details of data integration software, tools, platforms and vendors. A data-driven technique for a hybrid BDI using multilayer perceptron was discussed in this research [55]. A customized multilayer perceptron model was constructed using time-based parameters. The fields applied in optimization analysis are also used in the error matrix through additional neural network model. Research results revealed that this solution captures the variations in state variables. BDI project implementation for COVID-19 analytics was discussed [56]. This project was funded by the European Union research fund. This platform combines information from multiple sources such as world news, social media, published science and health data from healthcare institutions. The project design was co-created with industry, academia, health professionals, and policymakers to align with innovative technologies. This project successfully provides useful and actionable information to public health authorities.


**61**

**Authors**

Bansal Dhayne et al

Sazontev

2019

To develop a prototype of a big data

Useful for e-commerce data integration

domain

Data movement is faster than that of

Spark thus achieved optimization

Extensive details of big data integration

solutions for communities

integration system

et al

Chen et al Zheng et al

2015

To summarize categories and its

subcategories of data integration

techniques

Huang et al.

Portugal

2016

To perform spatial and temporal data

High-level representations by domain

specific languages, data analysis and

integration by model-driven techniques

analysis for assisting domain experts

et al

2014

To automate the data integration process

A more agile process for compelled

analysis by generating a subset of data

and leverage key capabilities

2015

To accomplish data integration of back-end

datasets in a complete manner

2019

To study healthcare data integration

Wide range of healthcare data

Data integration in

Y

Y

Y

Y

Y

Y

the healthcare sector

could not be done

efficiently using

traditional way

Lacks in methods for

Y

Y

N

N

N

Y

schema alignment

Integration of more

Y

Y

N

N

N

Y

Spark modules is not

supported

Since BDI methods

Y

Y

Y

Y

N

Y

behave differently in

different applications

it's difficult to select

the best data fusion

technique

HiperFuse modules

Y

Y

N

N

N

Y

are implemented

separately yet to be

integrated

provenance

Y

Y

Y

N

N

Y

technologies need

to be used in related

spatial–temporal

approaches

integration concepts, techniques and

tools are covered

methods, tools, and applications

2014

**Year**

**Study Objectives**

BDI by designing a Semantic Extract-

Transform-Load architecture

Publishing semantic data on the internet and thus contribute to the web of data

It is required to understand the heterogeneity of data i.e. ontology engineering

**Advantages**

**Disadvantages**

**1** Y

Y

Y

N

N

Y

**2**

**3**

**4**

**5**

**6**

*Big Data Integration Solutions in Organizations: A Domain-Specific Analysis*

*DOI: http://dx.doi.org/10.5772/intechopen.95800*


#### *Big Data Integration Solutions in Organizations: A Domain-Specific Analysis DOI: http://dx.doi.org/10.5772/intechopen.95800*

*Data Integrity and Quality*

**60**

**Authors** Fikri et al

2019

**Year**

**Study Objectives**

To get solutions for financial data real-time

integration issues and interpretation of

data

Cheng et al

Authors Bhandari

2020

et al

Vieira et al

2020

To conduct a literature survey of simulation

Simplification of the problem in the

absence of complex modeling

methods used for handling risks in the

supply chain with an emphasis on data

integration

Stonebraker

2018

To explore issues of BDI related to

Automation by machine learning and

Involves high cost

N

Y

Y

Y

N

Y

for domain experts,

shortage of training

data

Time complexity

Y

Y

N

Y

N

Y

is linear and it is

required to reduce

computation time

with an efficient

method

rule-based approach for augmenting

scalability in enterprises at Tamr region

et al

Ahmed et al

2016

To aggregate data from local and external

Classification accuracy improvement

and thus mitigates the impact of class

imbalance

resources, to generate mining table from

these, automatic generation of potential

discriminant features

Year

Study Objectives

To develop EJ screening, an adaptable

and community-based tool for the region

Houston Galveston Brazoria

Advantages Risk factor identification and

understanding among the communities

2020

To design BDI distributed architecture for

remote sensing data, where these data from

multiple sources

**Advantages** This real-time data integration solution

resolves earlier issues of classic ETL

tools

Improvement in performance with a

distributed architecture for remote

sensing data

**Disadvantages**

This solution cannot

Y

Y

Y

Y

N

Y

be integrated with a

hot production setting

Time and resource

Y

Y

Y

N

N

Y

complexity to

handle various preprocessing steps in

data integration

Disadvantages

reducing

environmental

disparities and

improving their

health and well-being

It is required to focus

N

Y

Y

Y

N

Y

on supply chain real

cases

1 Y

Y

N

N

N

Y

2

3

4

5

6

**1**

**2**

**3**

**4**

**5**

**6**


**Table 12.**

**63**

*Big Data Integration Solutions in Organizations: A Domain-Specific Analysis*

**2.12 Comparative analysis of survey papers with specific parameters**

ent operations which are performed on this heterogeneous Big data.

**4. BDI research issues, challenges and future directions**

implementation are discussed in the following sections.

real-time with low latency and high accuracy

**4.1 BDI research issues**

*The architecture of the BDI ecosystem.*

video

**Figure 5.**

The research issues, challenges and future directions related to BDI

1.Scalability - Scalable architectures for parallel big data processing

2.Real-time big data analytics - Stream big data processing of text, image, and

3.Deployment of the IoMT, IoT and CCTVs systems in smart environments would capture big data continuously. Processing multimedia big data in

Authors have selected fifteen BDI survey papers for comparative study. The specific feature parameters such as 1. Architecture, 2. Applications, 3. Open Issues and Challenges 4. Taxonomy, 5. Security, 6. Future Directions are used for comparing these survey papers. The Comparative analysis results are shown in **Table 12**.

The outline architecture of the BDI ecosystem is shown in **Figure 5**. This architecture has four major components. These components are Data Sources, Data Operations, Virtual Databases and Business Intelligence. This architecture also shows the operations performed by each of these components. The business Big data would be collected from various distributed sources in different formats and sizes in Data Sources component. The Data Operations component shows the differ-

The Big data gathered from various types of physically distributed databases are integrated to form a unified logical virtual database. Business intelligence information is extracted from this virtual data source by performing the operations stated in Business Intelligence Component. This intelligent information would be used for real-time intelligent business decision-making process across the organization.

*DOI: http://dx.doi.org/10.5772/intechopen.95800*

**3. The architecture of the BDI ecosystem**

*Comparative Analysis of curated survey papers with specific parameters.*

*Big Data Integration Solutions in Organizations: A Domain-Specific Analysis DOI: http://dx.doi.org/10.5772/intechopen.95800*
