Data Mining Techniques for Drug Use Research

This article discusses the use of data mining techniques in drug use research. It explores different data mining techniques and their application in analyzing drug substance. The article also highlights the importance of data mining in investigating drug addiction and supporting preventive measures.

Data Mining Techniques 0
Data mining techniques for drug use research
[Name]
[University Affiliation]
[Date]

Data Mining Techniques 1
1- Introduction
The literature review consists of 7
articles that has been gathered by using
Scopus and Google Scholar database. The
keywords that has been used while searching
for the relevant literature are “Data mining
Techniques and Drug Use”.
It has been identified in the research
that random cluster sampling of school has
been taken into consideration from the
Island of Mallorca from which 22 schools
were taken into consideration [3]. The
researcher has worked on random cluster
sampling; this method is used because the
researcher is able to divide the population
into different groups namely cluster. In case
of a random sampling of cluster method, it
can be said that using random cluster
sampling; the researcher is able to classify
the data into different groups [3]. Under this
research, 9300 students aged between 14
and 18 has been taken to analyse the
variables, cluster sampling method has
helped the study to look for unreliable
answers of adolescents, the final sampling is
based on 9284 adolescents (among which
47.1% were boys, and 52.9% were girls with
an average age of 15.59 years) [1][3].
2- Critical Analysis
For this type of study, researchers
are supposed to seek permission from the
participants; they even want written consent
to avoid any legal or ethical issue towards
their research [3]. Therefore, the researcher
took voluntary and written informed consent
from the participants to follow the protocol;
a qualitative research approach has been
taken into consideration. Data mining
techniques are of different types, for
example, for this research Decision tree and
Artificial Neural Network has been adopted
to carry out statistical techniques to look for
the specific types of drugs these children
have used in their lives [3] [7]. Other data
mining technique include ANN data
processing system to structure the data and

Data Mining Techniques 2
developed functioning of biological
networks to outline the characteristics of
parallel processing with distributed memory
to adopt the surrounding using
backpropagation algorithm in analysing the
data [3].
The results determined that
approximately 52.7% students have drunk
alcohol, 25% of them have smoked tobacco,
18.6% students use cannabis, and 1.6%
students use cocaine [3]. The research has
used the predictive power of motives for
drug use, therefore, data mining predictive
model has been taken into consideration,
Decision Tree has been run to take logistic
regression, K-Nearest Neighbour, Naïve
Bayes, and Artificial Neural Network to
analyse drug substance and see if there is
any balance within it [1][6].
In another research titled
“Employing data mining to explore
association rules in drug addicts” the
researchers have determined that in normal
terms data mining are only taken as the
technique to extract knowledge implicitly
from the existing database. The research
follows a process to analyse and study a
huge amount of data in order to extract
underlying facts and figure to determine a
pattern of drug use, and its relationship with
legal jurisdiction to increase the importance
of data mining techniques in investigating
descriptive and predictive purposes. The aim
behind this is to unveil the facts and figures
to extract the data in an understandable
manner [7]. The researcher has to use the
exploratory technique to identify the
variables and characteristics of the database
in order to predict any future detail
associated with this research. It has been
determined by the researcher that there are
different data mining technique to employee
require a database and its results to work
with various kinds of discovered knowledge
[6][7]

Data Mining Techniques 3
It has been identified that using
association rules with simple, clear, and
intuitive, intelligible with practical tools to
explore the relationship between variables
and descriptive ways to mine data in order to
identify the variables based on procedure
and processes that are efficient for data
collection and analysis [6][7]. The research
has followed all the rules associated with
data mining techniques, for example, it has
taken 57.96% single young people using
LSD, in other words, they are more likely to
involve in drug abuse. As for following rule
2, it determines that the drug abuse increases
the likelihood of rehabilitation 2.35 times.
Meanwhile, rule 3 determines that it have a
family history of drug abuse [7].
According to rule 4, young drug
addicts suffering are facing depression, they
are taking different types of drugs, for
example, LSD and Heroine to overcome
their depression, and in this situation, they
are more likely to commit a crime or get
involved in a dangerous situation [7]. The
research has outlined that in order to explore
drug addiction database, a researcher is
required to have knowledge about which
areas to use as it helps them look at the right
direction to help different social structures
and organisations [6][7]. It helps the
researcher to make the right decision about
drug addiction especially in supporting
preventive measures to plan rehabilitation
measures against drug addiction [6][7].
Graciela & Abeed in their research
have stated that before that data collection
method, it is important for a researcher to
work on the different set of drugs and
interests to work on associated comments
using social media [1][6]. For this
researcher, it has worked on different drugs
which are normally prescribed for chronic
diseases and syndromes to work on the
database of comments based on high
prevalence of use, for example as outlined in
IMS health Top 100 drugs used in 2013[1].

Data Mining Techniques 4
The initiation of the data mining process has
worked on initial drug list, Twitter has been
using to list down the drug names based on
its trade and generic keywords, for example,
it has worked on in-house pharmacology
expert to work on the backgrounds of drug
previously used in publication software [1].
In order to address the issue of
incorrectly spelt drug names influencing
data mining, it has worked on spelling
variation generator. The generator initially
distinguishes lexically close incorrect
spellings, explicitly those that are 1-alter
remove away as far as Levenshtein separate
[1]. It has worked on comparable incorrect
spellings are then recognised. Meanwhile,
the Google custom mining API is utilised to
distinguish a little arrangement of incorrect
spellings that are usually utilised by
participants. It has adopted a downloadable
adaptation to generate free accessible with a
self-loader generator [1][6].
The tweets seem to display various
kinds of data, for example, side
effects/signs, saw antagonistic medication
responses, prescription maltreatment data,
client slants towards medications and
additional costs, and potential medication
misuse, to give some examples [1][6]. This
has helped the researchers determine the
drug names regularly misspelt. It relies upon
the goal; particular sorts of medication-
related data can be mined from this data
source [1].
3- Difference in Point of Views
Qi, et al. has focused on the Proteins
as a primary element of drug use endured by
researchers and scientists from different
spaces [2]. It has been communicated with
each other to frame the premise of flag
transduction pathways and transcriptional
administrative systems. The researcher is
focused on using proteins of medication
targets to determine practical biomolecules
tending to constrained by some dynamic

Data Mining Techniques 5
mixes [2][3]. The research has gathered date
related to proteins from the Drug Bank
Database (Version 3.0) which has
approximately 1604 proteins to explain drug
targets. Further, the process of data mining
was forced to expel nonhuman proteins in
using groupings bigger than twenty per cent
by utilising PISCES system [2]. It is based
on the blend of iotas and atoms to determine
protein as a possibility for the medication
targets using variables such as water
dissolvability, hydrogen particle focus (pH),
bases attributes, and its structure [2].
Despite the fact, it is observed that
the association between databases to
conduct the screening process are not
dependable. This has been using for
different properties of proteins to work on
the fundamental compound or physical
properties of proteins generally [2]. The
research has selected specific properties in
the examination to work on essential
substance or physical properties of proteins.
It has been pursued the separating
procedure. There are some properties of
importance with the assignment were
separated, for example, peptide cleavages,
N-glycosylation, O-glycosylation, low
multifaceted nature locales, transmembrane
helices, to work on physical or compound
attributes. It can be said that these properties
were critical pieces of information in
choosing the natural action of proteins [2].
The research has utilised pepstats
which is online programming from
EMBOSS, to compute measurements of
properties. It has likewise called the
unlabelled proteins as unsure NDTPs
because of the previous earlier data about
the extents of DTPs in the dataset [2][3].
The unsure NDTPs were those as the
researcher did not know whether any of
them would be the medication target
hopefuls. At last, a gathered dataset with
517 known DTPs and 5376 unsure NDTPs
was utilised for the screening task. In

Data Mining Techniques 6
particular, a few proteins in the 5376
proteins would be suggested as in all
likelihood DTPs from the dataset of dubious
NDTPs [2][3].
Shadma, et al. outlined that learning
and knowledge extricated permits foreseeing
the conduct and future conduct of research
in order to shape its direction. It enables
entrepreneurs to take positive learning drive
choices. Data mining is implemented in
various space like FMCG, economy,
medicinal, instruction framework, and
others. This information has been extracted
from actuality by applying design
acknowledgement, factual, scientific
procedures those outcomes in skill type of
realities, patterns, affiliation, examples,
inconsistencies, and special cases [5]. There
are a few regions where data mining is
connected [5][6].
Data prospecting is an extraordinary
arrangement of consideration in the
educational media all in all as of late,
because of accessibility of mass data and up
and the coming requirement for
transforming such data into valuable data
and learning [5][6]. This is the method for
looking at important actuality and examples.
The hint picked up can be using amid
applications examination, for example,
misrepresentation identification, and client
maintenance, to generation stream and
scientific investigation [5]. Data prospecting
can be seen because of the normal
development of data innovation. It is an
iterative procedure [5][6].
It can be said that the data extracted
from the social chain is bound together in
order to deal with settling logical errands.
Since friendly chain contains plentiful of
data Author utilised Big Data standards to
mine and broke down it [5]. It is also found
that data from the amiable chain is mined
utilising the crawler or motor which spares it
into a group. Furthermore, the enormous
volume of mined data is sifted and

Data Mining Techniques 7
accumulated to get similarly modest datasets
of data that is appropriate to the explaining
task [5][6]. At last, gathered data is utilised
as a contribution for complex applications
which perform last and advanced data
examination.
Therefore, to oversee computational
method for the unpredictable application
Author utilised AaaS (Application as a
Service) demonstrate which is executed in
this condition for appropriated processing-
based cloud stage [5][6].
The author Ranjan has outlined in his
research that pharma companies are
following decision oriented, foundational
choice models to empower their
management in order to assess the result that
is relied upon to result from the usage a
selection program chosen according to the
nature of the research [4]. The model has
helped the study to follow the legal and
ethical protocol of research taken a large
number of people in increasing the
effectiveness of its authenticity [4][6].
Numerous associations work on intensive
research process before and after
introducing a new drug to understand its
efficiency among people as it help them
look for any issue in its results and the
information is use for further improvements.
The capacity to assemble an effective
prescient model relies upon past data [1].
4- Conclusion
Data Mining is intended to gain from
past progress and disappointments and will
almost certainly anticipate what will occur
straightaway (future expectation). One may
think why to use Data Mining in pharma
industry associations when the measurable
investigation has now been performed. The
Data Mining apparatus checks the factual
majority of the anticipated examples and
reports [4][6]. Data Mining can assist
research in looking for its required data from
a different database. Therefore, if data
mining tools look for 100% medications of

Data Mining Techniques 8
some specific expansive gathering have
included for the execution investigation [4]
[1]. However, among them, just ten drugs
have the attributes of superior evaluations, at
that point, the device can caution that it is in
all respects liable to be a mannerism of the
database instead of a standard prescient
example [4].
5- References
[1]
Graciela, G. & Abeed, S., 2016. Data, tools and resources for mining social media drug
chatter. Proceedings of the Fifth Workshop on Building and Evaluating Resources for
Biomedical Text Mining, p. 99–107.
[2] Qi, W., JinCai, H., YangHe, F. & JiaWei, F., 2017. Efficient Data Mining Algorithms for
Screening Potential Proteins of Drug Target. Mathematical Problems in Engineering.
[3] Rafael, J., Joella, A., Berta, C. & Elena, G., 2018 . Data mining techniques for drug use
research. Addict Behav Rep, Volume 8, p. 128–135..
[4] Ranjan, J., 2007. Applications of data mining techniques in pharmaceutical industry. Journal
of Theoretical & Applied Information Technology, 3(4).
[5] Shadma, Q., Sonal, R. & Shiv, K., 2017. Mining Social Media Data for Understanding Drugs
Usage. International Research Journal of Engineering and Technology (IRJET).
[6] Wu, X., Kumar, V., Quinlan, J.R., Ghosh, J., Yang, Q., Motoda, H., McLachlan, G.J., Ng, A.,
Liu, B., Philip, S.Y. and Zhou, Z.H., 2008. Top 10 algorithms in data mining. Knowledge
and information systems, 14(1), pp.1-37.
[7]] Zahedi, F. & Zare-Mirakabad, R. M., 2014. Employing data mining to explore association
rules in drug addicts. Journal of AI and Data Mining , 2(2), pp. 135 - 139.