Addis Ababa University

College of Natural Sciences

Department of Computer Science

Article Review on

A Data Warehouse Approach for

Dawit Zelalem (GSE/2662/15)

Advisor: Minale Ashagrie (PhD)

Addis Ababa, Ethiopia

July 2024
 Table of Contents
Introduction................................................................................................................................1

Discussions.................................................................................................................................3

Critiques.....................................................................................................................................4

Conclusion..................................................................................................................................5

Reference....................................................................................................................................5
Introduction

The data warehouse (DW) approach is a fundamental concept in the field of business
intelligence (BI), where a centralized, integrated, and subject-oriented repository of data is
used to support an organization's decision-making process. At the heart of this approach is the
consolidation of data from various sources, such as operational databases, enterprise
applications, and external data sources, into a single, unified repository, ensuring consistency,
accuracy, and usability through integration and transformation processes. The data warehouse
is designed to support specific business subjects or domains, rather than being organized
around the organization's operational systems, and it stores historical data, allowing users to
analyze trends and patterns over time, with the data being non-volatile and accumulating to
provide a stable and consistent data source for reporting and analysis.

The paper, presented at the 2019 IEEE 28th International Conference on Enabling
Technologies: Infrastructure for Collaborative Enterprises (WETICE), addresses the
challenge of integrating and analyzing spatiotemporal data in data warehouses (DWs) for
business intelligence, with a focus on the telecommunication industry. The paper addresses
the need for a DW schema that can effectively handle the complexity of temporal and spatial
data, which is essential for the industry's data analysis and reporting requirements.

The primary objective is to propose a DW schema that unifies temporal and spatial data to
enhance the understanding, querying, and management of such data for online analytical
processing (OLAP).

The methodology involves developing a starnest DW logical schema that efficiently

The specific objectives the article tries to achieve are: To propose a DW schema that
integrates temporal and spatial data into a unified framework for business intelligence, To
extend the starnest DW logical schema to support spatiotemporal data, enhancing the
understanding, querying, and management of such data for OLAP, To develop a case study
for the telecommunication industry, demonstrating the application and effectiveness of the
proposed schema through SQL queries, To implement the proposed schema in a prototype
DW, evaluating its performance and efficiency in handling spatiotemporal data, To contribute
to the field of business intelligence and data warehousing by providing a novel approach to

1
managing complex data types, which is particularly relevant to industries dealing with large
volumes of spatiotemporal data.

The data used in the article is quantitative, specifically call detail records (CDRs) from the
telecommunication industry. CDRs contain a variety of quantitative information such as call
duration, phone numbers involved in the call, time and date of the call, location data, and
other relevant metrics. This type of data is numerical and can be analyzed to uncover patterns
and trends, which is the focus of the proposed data warehouse (DW) schema in the paper. The
schema is designed to handle the volume and complexity of such quantitative data for
business intelligence purposes.

The method of data analysis in the article is based on the use of the proposed data warehouse
(DW) schema, which integrates temporal and spatial data. The analysis is performed through
structured query language (SQL) queries that are executed on the DW. These queries are
designed to extract meaningful information from the quantitative data stored in the DW,
which in this case includes call detail records (CDRs) from the telecommunication industry.

The paper does not go into detail about the specific algorithms or statistical methods used in
the analysis, but it implies that the analysis is facilitated by the structure of the DW schema,
which allows for efficient querying and processing of the data. The analysis is likely to
involve aggregation, filtering, and possibly some level of data mining to uncover patterns and
trends in the CDR data that are relevant for business intelligence purposes.

The implementation section of the paper mentions the use of Oracle Data Warehouse Builder
and SQL Developer for the prototype DW, which suggests that the data analysis is conducted
using standard relational database management tools and techniques. The performance of the
DW and the efficiency of the SQL queries are evaluated to assess the effectiveness of the
proposed schema in handling spatiotemporal data.

2
Discussions

The paper begins by discussing the importance of data warehouses (DWs) in the era of big
data and cloud computing, emphasizing the shift from traditional DWs to cloud-based DWs
with virtually unlimited storage. It highlights the role of DWs in business intelligence (BI)
and introduces the concept of online analytical processing (OLAP), explaining its relevance
to decision-making processes. The authors identify the challenge of managing and analyzing
spatiotemporal data, which involves both spatial and temporal dimensions, noting that
conventional multidimensional databases are inefficient for handling such data, especially in
industries like telecommunications where vast amounts of data are generated.

To address these challenges, the paper proposes a novel DW schema called the starnest
schema, an extension of the star and snowflake schemas designed to integrate temporal and
spatial data in a unified manner. This schema naturally represents hierarchical levels and
efficiently models aggregation levels for dimensions, and it extends to support spatiotemporal
data, allowing for the representation of both spatial and temporal dimensions within the DW.

A case study for a telecommunication company illustrates the practical application of the
proposed DW schema. The case study includes a fact table and four dimension tables, with
ten SQL queries demonstrating the schema's functionality and performance. The
implementation of the prototype DW using Oracle Data Warehouse Builder and SQL
Developer is detailed, including the hardware and software used, the size of the DW, and the
number of rows in each table.

The paper presents a novel approach to data warehousing that integrates spatial and temporal
aspects, rather than explicitly addressing competing theories. It highlights the potential of the
proposed starnest structure to more effectively and organically express hierarchical data by
drawing comparisons with traditional star and snowflake schemas.

The authors demonstrate the advantages of their proposed DW schema over traditional
schemas in order to assist with its implementation. They argue that the toughest schema can
more effectively represent the data's hierarchical structure and reduce the requirement for join
operations, which will enhance the efficiency of query processing. The argument is
reasonable because it is supported by the implementation outcomes addressed in the paper
along with the schema's logical structure.

3
The authors report on the performance of the DW, noting that queries using nested tables are
generally faster than those using flat tables. They discuss the implications of the proposed
schema for the telecommunication industry and its potential to improve decision-making
processes. The paper concludes by summarizing the contributions of the proposed DW
schema and its potential impact on the field of business intelligence and data warehousing.
Future work outlined by the authors includes testing the model with real data and across
different cloud vendors to generalize the solution for other business domains.

Critique

Strength

The paper well-structured and focused on a relevant problem in the field of business
intelligence and data warehousing. The proposed starnest schema is an innovative
approach that effectively addresses the complexity of spatiotemporal data. The case
study and SQL queries provide practical examples that validate the schema's utility
and performance.

 Clear focus on a relevant problem.

Weakness

The paper could benefit from a more detailed discussion on the limitations of the
proposed schema and potential challenges in its implementation. Additionally, while
the paper mentions future work with real data and across different cloud vendors, it
would be helpful to have more concrete plans for this research direction.

 Lack of comparison with other methods or theories.

4
Conclusions

The paper successfully proposes a novel DW schema for integrating and analyzing
spatiotemporal data in the context of business intelligence, with a specific application in the
telecommunication industry. The starnest schema is an effective solution that improves the
efficiency of OLAP queries by handling complex data dimensions. The case study and
implementation demonstrate the schema's practicality and potential for enhancing decision-
making processes in data-intensive industries.

Reference

[1] G. Garani, A. Chernov, I. Savvas, and M. Butakova, "A Data Warehouse Approach for
Business Intelligence," in 2019 IEEE 28th International Conference on Enabling
Technologies: Infrastructure for Collaborative Enterprises (WETICE), Napoli, Italy, 2019,
pp. 70-75, doi: 10.1109/WETICE.2019.00022.