3003 - BIGDATA - Nhóm 3

BIGDATA
Research trends on Big Data in Marketing:

A text mining and topic modelling based
literature analysis
GROUP 3 - PRESENTATION
BIGDATA Group 3 Presentation 2023
About Us
Group member
1. Trần Vũ – Leader
2. Nguyễn Năng Quốc Bảo
3. Nguyễn Thị Huỳnh Như
4. Nguyễn Thanh Thuỷ
5. Nguyễn Thị Kim Xuyến

CONTENT
Theory Model concepts
Method of Results and

Research discussion
Conclusion Contribution
THEORY
In this article, what theories are applied for research?
The article does not explicitly mention any specific theories or theoretical frameworks
applied for research. However, it does refer to some disciplines and techniques that are
related to Big Data, such as mathematics, statistics, optimization methods, signal
processing, data mining and machine learning, visualization approaches and social
network analysis.
BIGDATA Business Presentation 2020
Theory
Big data is getting

bigger
Big Data is everywhere these days, whether in the form of:
• Structured data, such as an organization's • Social networks such as Facebook and
traditional databases (e.g., CRM) Twitter are having a huge impact on
• Unstructured data, driven by new influencing customers’ decisions, leading
communication technologies and user organizations and brands to incorporate
editing platforms (e.g., text, images and information originated in such platforms in
videos) (Lansley & Longley, 2016). their Marketing solutions (Moro et al., 2016).
Theory
The 3 vs in big data

management:
volume, variety, and
velocity
(Laney, 2001)
Recently, two more Vs were included onto the
Big Data equation: Variability, and Value.
Theory
High volume, velocity and

variety information assets
cost-effective demand,
innovative forms of
information processing for
enhanced insight and decision
making (Fan & Bifet, 2013)
• In order to harness Big Data, researchers have developed new techniques to capture,
process, analyze and visualize large amounts of data within a limited processing
timeframe.
• These techniques involve various disciplines: mathematics, statistics, optimization
methods, signal processing, data mining and machine learning, visualization
approaches and social network analysis (Chen & Zhang, 2014).
• The technologies and tools for exploring data can be divided in three classes: batch
processing tools, streaming processing tools and interactive analysis tools.
Theory
A big data solution

typically includes the
four logical layers
(Mysore, Khupat, & Jain, 2013)
• Big Data sources – different channels from which the data is generated;
• Data messaging and store – layer responsible for acquiring and storing the data;
• Analysis – where the data and insights are extracted;
• Consumption – application of the analysis layer results.
Theory
A Big Data system

is only effective
for business when
it is set up for the
extraction of
useful knowledge
to support • Predictive analytics solutions can be used on top of Big Data to facilitate that purpose, in a combined
solution that can benefit from both the loads of data and advanced machine learning solutions (Hazen,
business decisions
Boone, Ezell, & Jones-Farmer, 2014).
• Predictive analytics based on Big Data drawn specifically for Marketing purposes are also called
Marketing analytics solutions, aiming to provide solid-ground Marketing understanding and techniques
(Sharda, Delen, & Tur.ban, 2015) for marketers to solve real-world Marketing problems (Grigsby, 2015).
Theory
Guidelines for traditional

customer relationship
management (CRM) solutions
to adopt for benefiting from
the Big Data explosion that
arose with social media
(Malthouse, Haenlein, Skiera, Wege, & Zhang, 2013)
• The prescription of Marketing analytics principles, in both the form of predictive models and adequate human
resources with analytical skills to take the most from CRM solutions.
• Other studies have also identified the need to incorporate analytics solutions in CRM encompassing widely known
concepts in Marketing.
• Such as the evaluation of customer lifetime value (Moro, Cortez, & Rita, 2015b).
Theory
Marketing analytic solutions

drawn on top of Big Data may
benefit organizations in a
wide range of problems.
• Identifying customers keener to respond positively to a telemarketing campaign (Moro, Cortez, & Rita, 2014).
• Constructing interactive reports and dashboards for managers or even unveiling interesting trends from what is
being said about the brand on social media (Lacoste, 2016).
• Big Data solutions may be considered the foundations for insightful systems that effectively support marketers,
relieving the burden of slow-paced human analysis (Sharda et al., 2015).

Theory
A gap was predicted for the

forthcoming years in terms of
Marketing enabling
technologies and Big Data
(Feinleib, 2012)
• The size of the gap is increasing at the velocity of the

inflation of Big Data availability (e.g., Erevelles, Fukawa, &
Swayne, 2016).
• The next years offer a relevant opportunity for research and
development of Marketing solutions that explore the
knowledge provided by Big Data.
MODEL CONCEPTS
ScienceDirect:
a recent time frame including the last 6 years, journal
articles (reviews and articles in the press)
Search:
following query was executed: (marketing OR branding
OR advertising OR “channel distribution” OR
Model
“consumer behavior”) AND (“big data” OR “data
concepts
analytics” OR “big dataset” OR “massive data” OR
terabyte OR exabyte OR zettabyte OR petabyte).
Dictionaries
• Big Data and Marketing
• Geographical
• Sectors to account for the main economic sectors
• Products dictionary, splitting literature into physical
goods and services.
Text mining
is a process of extracting useful information from
unstructured text data using techniques such as natural
language processing, information retrieval, and machine
learning. (Kao & Poteet, 2007)
Model
concepts
Document term matrix:
a mathematical matrix that describes the frequency of

terms that occur in a collection of documents. In a
document-term matrix, rows correspond to documents in
the collection and columns correspond to terms.
Wordcloud
is a data visualization technique used for representing
text data in which the size of each word indicates its
frequency or importance. Significant textual data points
can be highlighted using a word cloud. Word clouds are
widely used for analyzing data from social network
websites.
Model
concepts Latent Dirichlet allocation
is a generative statistical model that explains a set of

observations through unobserved groups, and each
group explains why some parts of the data are similar.
The LDA is an example of a topic model. In this,
observations (e.g., words) are collected into documents,
and each word's presence is attributable to one of the
document's topics. Each document will contain a small
number of topics.
METHOD OF
RESEARCH
• Literature selection
• Text mining and topic modeling procedure
Literature
selection
• This query includes both Big Data and Marketing
high-level concepts, considered equivalent and
relevant terms for each area (e.g.,Chen & Zhang, 2014 for
Big Data; Grönroos, 1989 for Marketing).
• The final number of articles collected was 1560,

from a total of 211 different journals, proving the
diversity of interested journals on Big Data in
Marketing.
Text mining and topic modeling procedure

Text
mining
and topic
modeling
procedure
RESULTS AND
DISCUSSION
• Terms Frequency
• Topic discovered
Terms Frequency Topic discovered

• The authors present and discuss the results of their text mining and topic
modeling analysis on the articles related to Big Data in Marketing.
• The authors identify and describe 12 topics that emerged from the analysis,
Most of the Authors are from
such as Big Data analytics, social media, customer behavior, cloud
the United States and
computing, etc.
China, next is Australia and
• The authors analyze the distribution and evolution of the topics over time
Spain. Africa does not have
and across different dimensions, such as geographic location, products, and
many Authors who
sectors.
contribute literature.
• The authors highlight the main trends and challenges in Big Data in
Marketing research, such as the gap between technological and business
domains, the lack of inter-continental collaboration, and the need for more
direct applications to business problems.
There are a large number of topics mainly characterized by three terms all related with Big Data (four topics, numbers 4, 9,
16 and 18, with a total of 236 articles) and with all three terms associated with Marketing (eight topics, numbers 2, 5, 6, 7, 8,
12, 14 and 17, with a total of 701 articles). Such a result means that from the total of 1560 articles collected, 937 (around
60%) are mostly focused on just one of the domains, limiting cross-disciplinary research.
Table 2 provides valuable insights on authors’ affiliations regarding geographic location. It clarifies if authors are producing
worldwide scale cross research on Big Data in Marketing, regarding distribution of continents. The results show that both
North American and European researchers are not developing research with co-authors from other continents, given that
both topics 1 and 3 achieve a value of zero for Europe and North America respectively, representing a perfect match to these
continents. The remaining two topics are closely related to Asia and Oceania, while also fitting articles published by South
American and African authors, respectively.
Table 3 presents the results of the LDA analysis when parameterized to five topics about Sectors. The first topic, with 963
articles (61.7%), is highly related to technology, which suggests that a large portion of researchers are pushing the edge on
technology related to Big Data. Next topic is mostly about consumer goods, while showing some vicinity to healthcare (beta
= 1.15). Topic number 3 provides the most solid inter-sectors relationship, between telecommunications and
industrials. The last two topics encompass fewer articles in comparison with the remaining and are mostly related to energy
and tourism.
CONCLUSION
• Research focus: Big Data application in Marketing

• Study period: 2010 - 2015
• Source: 1560 articles indexed in ScienceDirect's database
• Method: Text mining for efficient analysis
KEY FINDINGS
Increasing interest
in Big Data in
Marketing over the
year
Annex 2. Number of articles per year

KEY FINDINGS
Mostly from
Europe, North
America, and Asia
Few in South
Annex 3. Number of articles per
America and Africa Continent and year by Products
Annex 8. World map with authors’

affiliation frequency by country
KEY FINDINGS
Energy and
healthcare receive
half the attention
of consumer goods
in North America
Annex 8. Sectors frequency by continent

FOR FUTURE RESEARCH
Plenty of research is being conducted

on Big Data and Marketing
Specifying the benefits of Better alignment between Understand the main

Big Data toward Marketing technology and marketing requirements in designing
research Marketing solutions to
answer 5 Vs in Big data
CONTRIBUTION
• Theoretical contributions
• Practical Management Contributions
THEORETICAL CONTRIBUTION
01 02 03
Identifying Evaluate Theoretical
research and analyze contributions
trends on research
Big Data in trends
Marketing
01 Identifying research trends on Big Data in

Marketing 03
• Using text mining and topic modeling methods to analyze scientific articles related to
Big Data in the field of Marketing.
• Search for popular keywords, topics and the latest research directions in the field.
02 Evaluate and analyze research trends 03

• Using quantitative methods to assess researchers' interest in research topics related
to Big Data in Marketing.
• Analyze research trends over time, geography, and language and calculate what
percentage these trends account for in total articles.
03 Theoretical contributions 03
• Introduce new concepts, new research methods and new theories about Big Data in
Marketing.
• Provide new research directions and contribute to the building of theories and
models to analyze and explain the effectiveness of using Big Data in Marketing.
Building a
Marketing
Customers
Strategy based
on Big Data
Practical
Management
Contributions
Marketing New research
campaign directions
Building a Marketing Strategy

based on Big Data
• Based on the analyzed research concepts and trends,
Practical companies can identify new trends and find new
Management opportunities in the field of Marketing.
Contributions • Use analytical methods to collect and process data from

different sources such as social networks, e-commerce
websites to find information about customers, products,
services, etc.
Customers
• Apply analysis results to identify potential customers

Practical and devise appropriate marketing strategies to reach
Management them.
Contributions • Use analytics to uncover new customer trends and

tastes and create tailored products or services.
Marketing campaign
• Apply analysis results to evaluate the effectiveness of

Practical marketing campaigns, understand the characteristics of
Management participating customers, and make adjustments to

improve effectiveness.
Contributions
• Use analysis results to forecast market demand, helping
companies make appropriate strategic plans in
providing products or services.
New research directions
• Using analysis results to search for new research

Practical directions related to Big Data in the field of Marketing,
Management thereby releasing new scientific articles or improving

analytical methods to enhance efficiency and
Contributions effectiveness. accuracy of the analytical results.
Q&A
GROUP 3
THANKS FOR
LISTENING

3003 - BIGDATA - Nhóm 3

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3003 - BIGDATA - Nhóm 3

Uploaded by

Copyright:

Available Formats

BIGDATA

Research trends on Big Data in Marketing:

2. Nguyễn Năng Quốc Bảo

3. Nguyễn Thị Huỳnh Như

4. Nguyễn Thanh Thuỷ

5. Nguyễn Thị Kim Xuyến

Theory Model concepts

Method of Results and

In this article, what theories are applied for research?

related to Big Data, such as mathematics, statistics, optimization methods, signal

Big data is getting

• Structured data, such as an organization's • Social networks such as Facebook and

traditional databases (e.g., CRM) Twitter are having a huge impact on

• Unstructured data, driven by new influencing customers’ decisions, leading

communication technologies and user organizations and brands to incorporate

The 3 vs in big data

High volume, velocity and

A big data solution

A Big Data system

Guidelines for traditional

Marketing analytic solutions

being said about the brand on social media (Lacoste, 2016).

relieving the burden of slow-paced human analysis (Sharda et al., 2015).

A gap was predicted for the

• The size of the gap is increasing at the velocity of the

a mathematical matrix that describes the frequency of

is a generative statistical model that explains a set of

• The final number of articles collected was 1560,

Text mining and topic modeling procedure

Terms Frequency Topic discovered

• Research focus: Big Data application in Marketing

Annex 2. Number of articles per year

Annex 8. World map with authors’

Annex 8. Sectors frequency by continent

FOR FUTURE RESEARCH

Plenty of research is being conducted

Specifying the benefits of Better alignment between Understand the main

01 Identifying research trends on Big Data in

02 Evaluate and analyze research trends 03

Building a Marketing Strategy

Management opportunities in the field of Marketing.

Contributions • Use analytical methods to collect and process data from

• Apply analysis results to identify potential customers

Contributions • Use analytics to uncover new customer trends and

• Apply analysis results to evaluate the effectiveness of

Management participating customers, and make adjustments to

New research directions

• Using analysis results to search for new research

Management thereby releasing new scientific articles or improving

You might also like