i need help with my research seminars and methodology

Anonymous
timer Asked: Apr 1st, 2017
account_balance_wallet $25

Question Description

Literature Review Assignment

Review due date: 1 May, 2017.

Selection of topic/area/theme of review by: 20 March, 2017.

Submission instructions: Submit via moodle.

You are asked to write a Literature Review on a specific topic within one area of a Computer Science specialization that you are interested in (see example below on topic/CS area). Your review will be written in the form of a research paper according to the IEEE conference style format which can be found in: http://www.ieee.org/conferences_events/conferences/publishing/templates.html for Word and Latex.

The paper should be a minimum of 5 pages in the above format. (Latex is strongly recommended).

Instructions and advice:

  • Identify an area of specialization that you are interested in. Scan through all available sources to identify interesting, current research on a specific topic within the area. Students opting for the thesis option for their MSc, can use this literature survey to perform introductory-background work for their thesis.
  • After you have done some initial investigation, consult with a faculty member, who specializes in the area you are interested in. (If you do not know who to consult for a specific specialization/topic, ask me and I will advise you on who to contact).
  • With the help of the faculty member, you will clearly identify the topic and title of your review, as well as 3-4 relevant, recent journal or conference papers from where start.
  • Once you finalize the topic and theme of your review, submit them to me by March, 20 (as stated above), in order to have the review proposal approved.
  • Your review must include at least 8-10 papers published in refereed journals and conferences.
  • Your review should be organized thematically (see slides on Writing Research Papers, section on Literature Review). You do not report on each paper you read one by one. Your review is not a summary of studies, but a synthesis of information which requires comparing

themes, methods and conclusions among the different works. A good way of keeping track of all this work and organizing your review is with the use of a synthesis matrix (literature review matrix).

An example of such a matrix is shown below. This matrix is taken from the survey paper:

“Toward the Next Generation of Recommender Systems: A survey of the state-of-the-art and possible extensions”, by G. Adomavicius in IEEE Transactions on Knowledge and Data Engineering, 17 (6), 2005. (I uploaded it on moodle).


This survey is on Recommender Systems, which is a topic of Machine Learning (and Data Mining) which falls under the general area of Artificial Intelligence. The matrix shows how Recommender Systems can be categorized as: 1) content-based, collaborative or hybrid, based on the recommender approach used (rows) and 2) heuristic-based or model-based, based on the types of recommendation techniques used for the rating estimation (columns).

On moodle you will also find another two survey papers. I strongly recommend that you go through these survey papers to get an idea of the structure of literature surveys and how they are organized (they also include literature matrices).

Unformatted Attachment Preview

COMP-500 Research Seminars and Methodology Instructor: Dr. Athena Stassopoulou Literature Review Assignment Review due date: 1 May, 2017. Selection of topic/area/theme of review by: 20 March, 2017. Submission instructions: Submit via moodle. You are asked to write a Literature Review on a specific topic within one area of a Computer Science specialization that you are interested in (see example below on topic/CS area). Your review will be written in the form of a research paper according to the IEEE conference style format which can be found in: http://www.ieee.org/conferences_events/conferences/publishing/templates.html for Word and Latex. The paper should be a minimum of 5 pages in the above format. (Latex is strongly recommended). Instructions and advice: 1) Identify an area of specialization that you are interested in. Scan through all available sources to identify interesting, current research on a specific topic within the area. Students opting for the thesis option for their MSc, can use this literature survey to perform introductory-background work for their thesis. 2) After you have done some initial investigation, consult with a faculty member, who specializes in the area you are interested in. (If you do not know who to consult for a specific specialization/topic, ask me and I will advise you on who to contact). 3) With the help of the faculty member, you will clearly identify the topic and title of your review, as well as 3-4 relevant, recent journal or conference papers from where start. 4) Once you finalize the topic and theme of your review, submit them to me by March, 20 (as stated above), in order to have the review proposal approved. 5) Your review must include at least 8-10 papers published in refereed journals and conferences. 6) Your review should be organized thematically (see slides on Writing Research Papers, section on Literature Review). You do not report on each paper you read one by one. Your review is not a summary of studies, but a synthesis of information which requires comparing COMP-500 Research Seminars and Methodology Instructor: Dr. Athena Stassopoulou themes, methods and conclusions among the different works. A good way of keeping track of all this work and organizing your review is with the use of a synthesis matrix (literature review matrix). An example of such a matrix is shown below. This matrix is taken from the survey paper: “Toward the Next Generation of Recommender Systems: A survey of the state-of-the-art and possible extensions”, by G. Adomavicius in IEEE Transactions on Knowledge and Data Engineering, 17 (6), 2005. (I uploaded it on moodle). COMP-500 Research Seminars and Methodology Instructor: Dr. Athena Stassopoulou This survey is on Recommender Systems, which is a topic of Machine Learning (and Data Mining) which falls under the general area of Artificial Intelligence. The matrix shows how Recommender Systems can be categorized as: 1) content-based, collaborative or hybrid, based on the recommender approach used (rows) and 2) heuristic-based or model-based, based on the types of recommendation techniques used for the rating estimation (columns). On moodle you will also find another two survey papers. I strongly recommend that you go through these survey papers to get an idea of the structure of literature surveys and how they are organized (they also include literature matrices). ...
Purchase answer to see full attachment

Tutor Answer

Robert__F
School: Boston College

Good luck in your study and if you need any further help in your assignments, please let me know Can you please confirm if you have received the work? Once again, thanks for allowing me to help you R

Literature Review Assignment: Digital
Data Mining
Student name:
Course:
Tutor:
Date:

Data Mining

nontrivial extraction of implicit, useful
Digital

Data

mining

In

Event

track

information from the Digital Database and

intelligence

previously unknown. Thus it implies that we

Abstract

take only the most useful information that
was unknown to the organization and also

We live in the age of information, and often
which will play an important in the running
we have the belief that information is power
of the team (Witten, 2016).
and success. The advancement in technology
which has led to sophisticated computers,

Introduction

satellites and other critical technologies

Digital Data mining is a concept in the field

which have aided a great deal in collecting a

of computer science, and this idea is rising

tremendous amount of both the Digital Data

in its application because of the availability

and the information. The primary concern

of various information in the world. The

though is the kind of information we are

organization strives to get the most useful

receiving via cloud platform or mobility

information for their business while using

deveices

sophisticated

the minimal resources regarding the cost of

technologies in existence. It is at this point

getting that information and also when

that the concept of Digital Data mining

storing such information. Digital Data

comes where we screen and take only the

mining plays a significant role in screening

most important and the necessary type of

cloud

information

the

allowing the organization to get the useful

organization concern. Digital Data mining is

information about their organization. The

also called Knowledge-Discovery in Digital

researchers’ claims that we live in Digital

Databases (KDD), and it refers to as

Data age where information was generated

using

that

this

is

crucial

to

platform

information

and

only

1

Data Mining

at high speed, and thus the future of this

into a common source for further

generation

processing (Dong, 2017).

will

be

represented

with

gigabytes of Digital Data which will require

iii.

Digital Data selection: after all the

more storage space as well as the high

Digital Data are collected from

processing speed of the same Digital Data to

different

become useful information (Han, 2015).

relevant Digital Data are retrieved

This will help in flitering Digital Data by

for analysis. It is at this step that the

using annotated algorithem to filter the

organization decide what kind of

useful information requested in tracking

Digital Data that they want to

events at cloud platform.

retrieve from a pool of collected

and

grouped,

Digital Data.

Digital Data mining is an iterative process
which follows the following steps:

sources

iv.

Digital Data transformation: this
step is also known as Digital Data

i.

Digital Data cleaning: this is the
consolidation and it is a phase where
most important step in Digital Data
the selected Digital Data from the
mining where noise Digital Data and
previous steps are changed into
irrelevant Digital Data are removed
forms that are appropriate and useful
from the Digital Data collecting in
in the mining procedure (Oliveira,
the field or from other sources.
2012).

ii.

Digital Data integration: after the
v.

Digital Data mining: this is the most

Digital Data is cleaned, multiple
important step in the entire process
Digital Data sources which are often
of Digital Data mining where the
heterogeneous are grouped together
clever techniques are applied so as to

2

Data Mining

extract useful patterns from the
transformed Digital Data.
vi.

Data
transformation

Pattern evaluation: once the patterns
are extracted, the interesting patterns
which

represent

knowledge

are
Data mining

identified using the given measure.
vii.

Knowledge representation: this the
last step in the Digital Data mining
process where the identified or
discovered

knowledge

from

Pattern
evaluation

the

entire process is presented to the end
users. Visualization techniques find
its place here since it helps to
understand and interpret the result of
Knowledge
representation

the Digital Data mining.

The entire process of Digital Data
mining can be represented as below:

Source of Digital Data to be mined
There are many sources where the

Data
cleaning

Digital Data to be extracted are
retrieved,

and

some

examples

include:
Data
integration

i.

Flat files: this one of the most

common sources of evidence used in
Digital Data mining algorithms at the

Data selection

research level. Flat files are Digital
Data files in a binary format which
3

Data Mining

contain Digital Data which can either

regarding

be time-series Digital Data, scientific

Digital Data collection (Kimball,

measurements, transactions or more

2011).

(DuMouchel, 2009, August).
ii.

iv.

resources

needed

for

Transaction Digital Database:

Relational Digital Databases:

this is the source of sets of records

Digital Data mining algorithm is

which represent the operations. Each

more efficient when it comes to

record has a time stamp, an identifier

relational Digital Databases since it

and a set of items. This is the source

will take advantage of Digital Data

of descriptive Digital Data for

selection,

different items which might be

consolidation,

and

transformation. In addition to this
features,

SQL

goes

further

by

allowing the element of prediction,
deviations, detection among others.
iii.

useful to the organization.
v.

Multimedia Digital Database:

this includes Digital Data collected
from audio, video, images and text

Digital Data warehouse: this

media. This source makes the entire

represents a repository of Digital

process of Digital Data mining more

Data which has been collected from

challenging because of the way

many sources. This is a primary

Digital Data is represented with high

source of various Digital Data since

dimensionality. This source calls for

it originates from multiple sources. It

computer vision, computer graphics,

allows for analyzing Digital Data

natural language processing and

from multiple sources under the

image interpretation methodologies

same roof and thus less expensive
4

Data Mining

so as to ease the entire process of
pattern extraction from this sources.
The following figure shows the
whole

process

of

Digital

Data

mining. It is extracted from Digital
Data

Mining

Concepts

and

Techniques by Jiawei Han (Kamber,
2011). The following figure shows
the entire process of Digital Data
mining. It is extracted from Digital
Data

Mining

concepts

and

techniques by Jiawei Han (Kamber,
2011).

Digital Data mining functionalities
i.

Characterization: this concept means the

summarization of the general features of the
objects in the target class and the result of
this function is specific rules. In this case,
the paramount Digital Data to any userspecified class is retrieved using a Digital

5

Data Mining

Database query and summarized so as to

iv.

yield importance of Digital Data at different

classification is established, the class label

levels. The mostly used methods here is

of different objects can be predicted using

attribute-oriented induction method.

the attribute value of the objects and those of

ii.

Discrimination: this is the comparison

of the general features of objects, and this is
between two classes which are frequently
referred to as target class and different class.

Prediction: once a model for

the class as well. The central concept here is
the use of the current values so as to
determine or predict the most probable
future values.

The result of this process is discriminant

v.

rules and the methods used in the process is

classification as it leads to the grouping of

the same with those of the characterization.

Digital Data into different classes. The

iii.

Classification: this is the arrangement

of Digital Data into different given classes.
This step makes use of quality levels to put
different

objects

in

the

Digital

Data

collection. A training set is used which
group different objects associated with the
known class labels. The algorithm used for
classification learns from the training set and
builds a model which is used in the
classification of the new objects (Matijevič,
2014).

Clustering: this is the same as

significant difference with classification is
that in clustering, class labels are unknown
and it is the prime responsibility of the
algorithm being used to determine the
acceptable level to place the objects
(Cacciari, 2010). This process is also called
unsupervised

classification

since

the

classification is not dependent on any given
class labels. Most of the methods or
approaches used under clustering include
intra-class

similarity

and

inter-class

similarity.

6

Data Mining

vi.

Outlier analysis: this is exceptions

responses on different available

which cannot be grouped in any known
class. Outliers are important in the analysis

products.
iii.

as it serves to show the domains.

in

different

the

customer

groups: as it helps in marketing
campaigns

Digital Data mining is the most appropriate
concept

Determining

organization

it

also

helps

in

coming up with customer groups

and

using the surveys. The surveys

business and has played a key role in

used are forms of Digital Data

bringing many benefits to the organizations.

mining where Digital Data on
unknown products and services
Benefits of Digital Data mining
i.

ii.

Identification

of

are collected.
shopping

iv.

Banking: Digital Data mining

patterns: often times when trying

gives useful information when it

to get a suitable shopping pattern,

comes to loan information and

there are many issues and this

credit. It uses a model built from

issues can be solved using the

historical Digital Data of the

Digital Data mining since this

customers,

method helps in identification of

financial institution so as to

all the shopping patterns.

determine good or bad loans. It

Marketing campaigns: Digital

also helps the bank to identify

Data mining is beneficial when it

fraudulent credit card transaction

comes to marketing campaigns as

and thus useful in protecting the

it helps in identifying customer

credit card’s owners.

the

bank,

and

7

Data Mining

v.

Manufacturing:

Digital

Data

information

wrongly.

The

mining plays an important role

information might get into the

when it comes to manufacturing

hands of unethical people in the

sector

organization and thus its misuse.

since

manufacturers
faulty

it

helps

to

equipment’s

determine
and

Conclusion

thus
Digital Data mining is the most useful

determine

optimal

control
techniques especially in the business

parameters.
environment as it gathers the most useful
information

that

will

boost

its

operations. Digital Data mining is the
Disadvantages of Digital Data mining
i.

best technique for the purposes of

Privacy issues: there is no

resource

management

and

the

personal privacy since Digital

profitability of the business organization.

Data need to be analyzed. The

The issue of security is the arising issue

information

collected

from

where the organization tries to protect

customers

concerning

their

their Digital Data and information and

purchasing

trends,

credit

there is need to keep the collected

information

among

other

Digital Data and information from

information is exposed and this

unethical people.

makes the entire process of being
hard.
ii.

Misuse of information: there is a

References
Cacciari, M. S. (2010). The anti-kt jet clustering
algorithm. Journal of High Energy
Physics.

tendency of using the collected
8

Data Mining
Dong, X. L. (2017). Big Digital Data integration.
In Digital Data Engineering (ICDE), 2013
IEEE 29th International Conference.
DuMouchel, W. V. (2009, August). Squashing
flat files flatter. In Proceedings of the
fifth ACM SIGKDD international
conference on Knowledge discovery and
Digital Data mining.
Han, J. P. (2015). Digital Data mining: concepts
and techniques. Elsevier.
Kamber, M. H. (2011). Digital Data mining:
Concepts and techniques. Elsevier.
Kimball, R. &. (2011). the Digital Data
Warehouse? ETL Toolkit: Practical
Techniques for Extracting, Cleaning,
Conforming, and Delivering Digital
Data. John Wiley & Sons.
Matijevič, G. P. (2014). Kepler eclipsing binary
stars. III. Classification of Kepler
eclipsing binary light curves with locally
linear embedding. The Astronomical
Journal.
Oliveira, S. R. (2012). Privacy preserving
clustering by Digital Data
transformation. Journal of Information
and Digital Data Management.
Witten, I. H. (2016). Digital Data Mining:
Practical machine learning tools and
techniques. Morgan Kaufmann.

9

pl...

flag Report DMCA
Review

Anonymous
awesome work thanks

Similar Questions
Related Tags

Brown University





1271 Tutors

California Institute of Technology




2131 Tutors

Carnegie Mellon University




982 Tutors

Columbia University





1256 Tutors

Dartmouth University





2113 Tutors

Emory University





2279 Tutors

Harvard University





599 Tutors

Massachusetts Institute of Technology



2319 Tutors

New York University





1645 Tutors

Notre Dam University





1911 Tutors

Oklahoma University





2122 Tutors

Pennsylvania State University





932 Tutors

Princeton University





1211 Tutors

Stanford University





983 Tutors

University of California





1282 Tutors

Oxford University





123 Tutors

Yale University





2325 Tutors