01 - Overview Process

8/2/2019 01 - Overview Process

1/37

Research Overview and ProcessAssoc. Prof. Dr. Shamsul Anuar Mokhtar


2/37

What is research?

As a general term,research is gathering

information thatanswers a question andso solves a problem.(from Booth, Colomb,

& Williams The Craftof Research)

MIIT Universiti Kuala Lumpur AP Dr. Shamsul Anuar Mokhtar


3/37

What is research?

eg journalist finds outwho contributed toelection campaignfund, to understandpolitical decisions

eg repairer finds out

what procedure to useto fix a washingmachine



4/37

Academic Research

In academic research,you must not onlyanswer a question, but

you must findsomething new and

interesting.



5/37

Academic Research

You join a community of researchers.

You must advance the collective understanding of

this community.

Each community has a cumulative tradition with a set ofinteresting questions, tools and methods, practices, a

style and language for writing up the research.

Research is a conversation and ongoing socialactivity!



6/37

Academic Research

You need critical and careful reading of published

research

to learn what the community already knows

to fit your work into the community

to be prepared for your own work to be evaluated



7/37

Benefits of research project

The development ofskills to function more

independently

The opportunity to putclassroom knowledgeinto practice

The identification ofcareer interests

AP Dr. Shamsul Anuar MokhtarMIIT Universiti Kuala Lumpur


8/37


The building of mentorrelationships betweenfaculty and students

The stimulation thatcomes with criticalthinking

The opportunity toparticipate in newdiscoveries



9/37


The ability to betterunderstand researchmethodologies

The stimulation ofcreativity

The increased likelihoodof acceptance intograduate or professionalschool



10/37


The opportunity toearn academic credit

The sheer excitementcreated by intellectualactivity

The enhanced abilitygrasp the philosophy oflife-long learning



11/37

Types of research project

Independent systems

Integrated systems

Computational andmodelling project

Evaluation project



12/37

Independent systems

Independent systemsare IT productdevelopments whichlead to full

applications.



13/37

Integrated systems

Developing integrated

systems involves thedesign andimplementation of asubsystem which adds

to or extends workdone by others.



14/37

Computation & modelling

Computation andmodelling projects donot have as their primaryconcern the productionof IT products, but maydemonstrateunderstanding and

achievement via e.g.,prototyping, proof ofconcept product, etc.



15/37

Evaluation

Evaluation projects alsodo not set out toproduce IT product astheir primary aim,though again this doesnot rule out theproduction of IT

product whichsubstantiates claimsmade in the report.



16/37

A Question

Every piece of researchshould address aquestion of interest tothe community



17/37

A Question

Each community has a tradition of style of question,

what happens?

why does it happen?

how should one do something?

what something should one do?



18/37

A Question

Many questions fit into an on-going agenda

eg find data models to represent different sorts ofinformation

eg move data and computation in a network to

exploit locality

eg understand the design process



19/37

A Claim

Every piece of researchmakes a claim (the

contribution)thisshould answer aquestion of interest

Claims can be verydiverse, among fieldsand within fields



20/37

A Claim

This is what happens

eg how often is data corrupted when using weakconcurrency control

This is why something happens

eg what factors lead to project success in open-source development



21/37

A Claim

This is a better way to do something

eg efficiently recalculate a graph layout after achange to topology

This is a better something to do

eg allow users to see the model of their skills kept ina teaching system


Be explicit about themeaning of better


22/37

Evidence

You must back up theclaim

Each research method isdefined by the sort ofevidence that it canproduce

Each community has itsown standards of qualityand reasonableness



23/37

Evidence

Evidence can be very varied, for examples

a prototype implementation to show that a systemcan be built to achieve claimed functionality

a simulation model which is executed to show asystem has certain properties



24/37

Evidence

Evidence can be very varied, for examples

measurements of a running system to show it hasgood performance

observations of behaviour in an organisation toshow what is happening

a mathematical proof to show that some processhas desired properties



25/37

Argument

You should show thatthe evidence you offersupports the claim youmade



26/37

Argument

Its essential that youdeal with natural orobvious objections tothe correctness orimportance of the work

You must think like

your readers, andanticipate theirreactions



27/37

Argument

In systems work, this isoften called anevaluation of thedesign



28/37

Claim/Argument - Type 1

This system design leads to better performance on some metric

make sure you limit how much worse this makes other

metrics (such as cost!)

make sure your measurements are fair (dont compare withstrawman design but with state-of-the-art)

This system design off

ers better functionality for some uses

make sure you show it can be implemented with adequateperformance



29/37

Claim/Argument - Type 2

This behaviour can be explained by this theory

make sure you dont have confounding factors such

as level of experience, or method novelty, or subjectexpectations (placebo effect)

This is what happens

make sure you dont interfere too much with whathappens when you gather data, or misinterperete itdue to observer expectations



30/37

Common Mistakes 1

Gather lots of datawithout a focussedquestion or method

A collection of facts isnot a contribution!

it must reveal some

pattern orunderstanding thatyou make explicit



31/37

Common Mistakes 2

Build a system without afocused question orplanned evaluation

eg lets see how touse aspect-orientedprogramming in asensor network

eg let design agentsbreed and evolve



32/37

Common Mistakes 2

An innovative system is notnecessarily a contribution!

It must be a worthwhileinnovation in a sense youmake explicit

eg better functionality

eg better performance

eg better model of reality



33/37

Negative Results

Sometimes, you dontget the result youhoped for



34/37

Negative Results

You gather data that does not reveal any pattern orunderstanding

eg no factor seems to correlate well with projectsuccess

You design a system that turns out to be worse than thestate-of-the- art

eg your machine learning algorithm runs slowerthan expected



35/37

Negative Results

You can still salvage a thesis

but publication will be hard, and it wont have much

impact

Try to find some way to contribute to our understanding, orsuggest fruitful directions for further work

eg what features of the algorithm make it slow

Make sure the problem is intrinsic, not just your bad coding/experiment design/etc



36/37

Disclaimer

These presentation slides are designed, organised andcompiled by AP Dr. Shamsul Anuar Mokhtar based in partsfrom materials by A. Fekete, J. Davis, and others. The images

are downloaded from the World Wide Web. Copyright ofthe content and images are owned by the respective authors.

These presentation slides shall be used solely for sharing ofknowledge and reference purposes only. They shall not to betransferred to others outside the university nor shall they bepublished to the World Wide Web except with the consentof the author.



37/37

Research ProposalAssoc. Prof. Dr. Shamsul Anuar Mokhtar

Next

01 - Overview Process

Documents