RESEARCH PAPER WRITING

A Complete Guide - From Choosing a Topic to Getting Published

Transcript

1. RESEARCH PAPER WRITING A Complete Guide - From Choosing a Topic to Getting Published By – Shikhar Parteek Pandey Assistant Professor Department Of CSE GL Bajaj Institute of Technology & Management
2. WORKSHOP ROADMAP What We Will Cover Today 01 02 03 Foundations Where to Publish Anatomy of a Paper Choosing a topic, Google Scholar & search tools Journals, conferences, indexing & impact factor Title, abstract, introduction, contributions 04 05 06 Building the Core Results & Wrap-up Practical Toolkit Literature review, methodology, metrics Ablation study, comparison, conclusion, references Free tools, common mistakes, submission checklist Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 1
3. GETTING STARTED Why Write a Research Paper? 1 Builds Deep Expertise Forces rigorous, structured thinking about a problem in your domain 2 Strengthens Your CV Essential for higher studies (M.Tech/PhD), placements & scholarships 3 “Publish or Perish” A research paper is proof that you can identify a real problem, study existing solutions, propose something new, and validate it with evidence. Contributes to Knowledge Adds a citable, verifiable result to your field for others to build on 4 Personal Branding Remember: A good paper always tells a story — Problem → Gap → Idea → Proof. Google Scholar / Scopus author profile & citation record open doors Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 2
4. STEP 1 How to Choose a Research Topic 1 Pick a Domain Start from a subject you enjoy: AI/ML, IoT, Cybersecurity, Networks, etc. 2 Scan Recent Trends Browse the last 3-5 years of papers on Google Scholar / Scopus in that domain 3 Spot Recurring Gaps Note limitations authors themselves mention in their 'Future Work' sections 4 Check Feasibility Confirm dataset availability, tools, compute power & timeline 5 Discuss With Your Guide Validate scope, novelty and depth with your supervisor 6 Confirm Novelty Re-search the exact idea to be sure it isn't already solved Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 3
5. STEP 2 Mastering Google Scholar scholar.google.com Use precise search phrases Set up Alerts Put exact phrases in quotes, e.g. "feature selection" "intrusion detection" Get emailed when new papers matching your query are indexed Filter by year Check author h-index Use the left-side year filter to stay within the last 3-5 years Open an author profile to gauge research impact & activity Follow "Cited by" Export citations Click Cited-by to find newer papers building on a key reference Use the quote icon to export in APA / MLA / BibTeX instantly Use "Related articles" Quickly discover similar work you may have missed Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 4
6. STEP 2 (CONTD.) Other Literature Search & Reference Tools SC Scopus Elsevier's abstract & citation database — check journal indexing & CiteScore CO Connected Papers Visual graph of related papers to map out a research area quickly Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management WE Web of Science Clarivate's database — home of SCI / SCIE / ESCI listings & Journal Impact Factor RE ResearchGate Connect with authors directly; request full-text of paywalled papers SE Semantic Scholar AI-powered search with influential-citation highlighting & TL;DR summaries IE IEEE Xplore Primary source for IEEE conference & transaction papers 5
7. WHERE TO PUBLISH The Publication-House Landscape Publication House IEEE Springer Elsevier Wiley Taylor & Francis / CRC IEEE Xplore → Conferences (Scopus) LNNS, AISC series — not every series is Scopus ScienceDirect journals ✓ SSRN preprints ✗ (not peerreviewed) Peer-reviewed journals → Scopus indexed Journals & books → Scopus indexed Tip: Always verify a journal / series' current indexing status directly on the Scopus or Web of Science site before submitting — indexing can change year to year. Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 6
8. WHERE TO PUBLISH (CONTD.) Understanding Journal Indexing SCI / SCIE Scopus ESCI Non-indexed / Preprint Web of Science core collection — the most selective, highest-prestige tier Elsevier's large abstract & citation database; broadly recognised for research evaluation worldwide Emerging Sources Citation Index — newer / developing journals on their way to full SCIE status e.g. SSRN preprints — useful for early visibility but NOT peer-reviewed or citation-counted Best of all: IEEE / ACM Transactions — top journals of their respective societies, almost always SCI + Scopus indexed and highly cited. Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 7
9. WHERE TO PUBLISH (CONTD.) What Is Impact Factor (IF)? Impact Factor measures how frequently a journal's articles are cited, on average, in a given year. Journal Impact Factor (Year Y) = Higher IF ≠ automatically the right journal — check scope fit first IF is discipline-dependent: compare journals only within the same field Citations in Y to articles published in (Y-1, Y-2) Citable items published in (Y-1, Y-2) Also check the Quartile (Q1 highest – Q4 lowest) within its subject category Published annually by Clarivate in the Journal Citation Reports (JCR) CiteScore (Scopus) is a similar, parallel metric from Elsevier Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 8
10. WHERE TO PUBLISH (CONTD.) Conference vs. Journal vs. Book Chapter Conference Paper Journal Paper Book Chapter Review Speed Fast (weeks) Slow (months) Moderate Review Rigor Moderate High / multi-round Moderate Typical Length 4-6 pages 8-15+ pages 15-25 pages Indexing Often Scopus (e.g. IEEE Xplore) SCI / SCIE / Scopus / ESCI Scopus or Non-Scopus Best For Early results, quick visibility Higher studies, strongest credibility Broader survey / thematic work Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 9
11. PAPER ANATOMY The Structure of a Research Paper — At a Glance 1 Title 2 Abstract 3 Keywords 4 1. Introduction 5 2. Literature Review 6 3. Proposed Methodology 7 4. Experiments & Results 8 5. Conclusion 9 References We will walk through each of these sections in detail — what to write, how much to write, and common pitfalls. Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 10
12. PAPER ANATOMY (CONTD.) Title, Abstract & Keywords T Title A Abstract K Keywords 150-250 words covering: • Concise, specific, keyword-rich • Background (1-2 lines) • 5-6 terms, comma separated • Avoid vague words like "A Study on..." • Problem / gap addressed • Mix broad + specific terms • State the method + application area • Proposed method (1-2 lines) • Include method & domain names • Ideally under 15 words • Key results with numbers • Improves discoverability & indexing • One-line significance Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 11
13. PAPER ANATOMY (CONTD.) Section 1 — Introduction Intro Traditional Approach Problem Solution Keep it unbalanced on purpose: spend fewer words recapping the traditional approach, and more words on the Problem and your Solution. 1.1 Motivation (180 – 250 words) 1 What is the problem in the existing environment / current systems? 2 How do you propose to solve it — in one or two motivating sentences? Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 12
14. PAPER ANATOMY (CONTD.) Section 1.2 — Stating Your Contributions i Novel Technique e.g. a new feature-selection method — what makes your approach original ii Data Handling Preprocessing & normalization steps applied before modelling GA + iii Qual. & Quant. Analysis Accuracy, Precision, Recall, F1-score, Specificity + error-rate metrics PSO + iv Ablation Study Proof that each component of your model contributes meaningfully Bi-LSTM + v Benchmarking Proposed model vs. existing results — showing superior performance Example: Hybrid Model CNN = Novel Combined Architecture Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 13
15. PAPER ANATOMY (CONTD.) Performance Metrics You Must Report Metric What It Measures Accuracy Overall correctness of the model Precision Of predicted positives, how many are correct TP / (TP + FP) Recall (TPR) Of actual positives, how many are found TP / (TP + FN) F1-Score Harmonic mean of Precision & Recall 2 × (P × R) / (P + R) Specificity (TNR) Of actual negatives, how many correctly identified FPR / FNR False Positive / Negative Rate FP/(FP+TN) · FN/(FN+TP) FDR / FOR False Discovery / Omission Rate FP/(FP+TP) · FN/(FN+TN) Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management Formula (TP/TN/FP/FN) (TP + TN) / (TP + TN + FP + FN) TN / (TN + FP) 14
16. BUILDING THE CORE Section 2 — Writing the Literature Review 1000-1500 15-18 Recency Matters words papers referenced, 2021-2025 Prioritise the most recent 3-5 years of work Organise every paper you review in a comparison table: Ref Proposed Method Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management Dataset Finding Pros Cons Result 15
17. BUILDING THE CORE (CONTD.) Identifying Research Gaps & Problem Statement From Literature Review to Gaps 1 2 3 4 Compare limitations across the 15-18 papers reviewed Close the Loop For every research gap identified, explicitly state how your proposed model addresses it: Gap 1 Poor accuracy on imbalanced data → addressed by Proposed Model Gap 2 High false-positive rate → addressed by Proposed Model Gap 3 No feature-selection strategy → addressed by Proposed Model Cluster similar limitations into 5-7 distinct research gaps Write a crisp Problem Statement (2-3 sentences) Ensure every gap is specific, evidenced, and addressable Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 16
18. BUILDING THE CORE (CONTD.) Section 3 — Proposed Methodology 3.1 Overview One diagram + short paragraph summarising the end-to-end pipeline 3.2 Data Preprocessing Cleaning, handling missing values, encoding, train/test split 3.3 Feature Selection Technique used and why it was chosen over alternatives 3.4 Proposed Architecture Block diagram of your model, with clearly labelled layers/stages 3.5 Algorithm Formal pseudocode listing the step-by-step procedure 3.6 Hyperparameter Tuning Table of hyperparameters explored & final chosen values Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 17
19. BUILDING THE CORE (CONTD.) Section 4 — Experimental Setup & Result Analysis 4.1 Dataset Description Source, size, class distribution, train/test/validation split ratio 4.2 Performance Metrics Define each metric formally; benchmark against 8-10 recent papers 4.3.1 Qualitative Analysis Confusion matrix, ROC curve, loss/accuracy curves — visual evidence 4.3.2 Quantitative Analysis Numeric result tables with statistical comparison across metrics Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 18
20. BUILDING THE CORE (CONTD.) Section 4.4 — Ablation Study Systematically remove one component at a time to prove each part of your hybrid model earns its place. Case Configuration Accuracy Precision Recall F1-Score A Proposed Model (full) Highest Highest Highest Highest B Without Bi-LSTM ↓ ↓ ↓ ↓ C Without PSO ↓ ↓ ↓ ↓ D Without CNN ↓ ↓ ↓ ↓ E Without GA ↓ ↓ ↓ ↓ Case A is your full proposed model. Each subsequent row removes one module (Bi-LSTM, PSO, CNN, GA) — showing performance drop and proving its necessity. Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 19
21. BUILDING THE CORE (CONTD.) Section 4.5 — Proposed Method vs. State of the Art Build a table AND a chart comparing your model's accuracy against 8-10 recent benchmark papers on the same / similar dataset. Illustrative Comparison — Replace with Your Actual Results 99 99 97 96 95 95 93 93 91 91 89 87 85 SOTA-1 (2022) SOTA-2 (2023) Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management SOTA-3 (2024) SOTA-4 (2024) Proposed Model 20
22. WRAPPING UP Section 5 — Conclusion, Future Work & Societal Impact Conclusion 1 Summarise the problem, your proposed solution and headline results in 3-5 sentences — no new information here Future Research Directions 2 Honestly note current limitations and 2-3 concrete directions other researchers could pursue next Societal Impact 3 One short paragraph on real-world relevance — who benefits and how, beyond the numbers Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 21
23. WRAPPING UP (CONTD.) References & Overall Paper Length Citation Style Typical Length Guide (IEEE format) IEEE numbered style [1], [2] — most common for engineering papers Abstract 150-250 words APA (author-date) — common in Springer / social-science-adjacent venues Introduction + Motivation 500-800 words Always follow the target journal's own template & style guide Literature Review Use a reference manager (Zotero / Mendeley) to auto-format citations Methodology + Results Reference count: typically matches your literature review (15-25+) Overall Paper Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management 1000-1500 words (15-18 papers) 1500-2500 words ~4000-6000 words / 8-12 pages 22
24. PRACTICAL TOOLKIT Free Tools for Research Paper Writing ✓ Writing & Formatting Overleaf (LaTeX), Google Docs, Grammarly (free tier) ✓ Plagiarism Check Quetext / Small SEO Tools (free tier), institute's Turnitin access Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management ✓ Reference Management Zotero, Mendeley — auto-generate citations & bibliographies ✓ Data, Code & Stats Google Colab, Jupyter Notebook, Python/R, Weka ✓ Literature Search Google Scholar, Semantic Scholar, Connected Papers, ResearchRabbit ✓ Diagrams & Charts draw.io (diagrams.net), Canva (free tier), Excel/Origin 23
25. PRACTICAL TOOLKIT (CONTD.) Common Mistakes to Avoid ✕ ✕ ✕ ✕ Plagiarism or close paraphrasing without proper citation No clearly stated research gap — paper feels like 'yet another model' Poor English / grammar — always proofread or use Grammarly Overclaiming contributions beyond what results actually show Research Paper Writing Workshop | GL Bajaj Institute of Technology & Management ✕ ✕ ✕ ✕ Weak or outdated literature review (missing last 2-3 years) Skipping the ablation study — reviewers will ask for it Ignoring the target journal's formatting template Submitting to non-indexed / predatory journals without checking Scopus/WoS first 24
26. SUBMISSION CHECKLIST ✓ Topic finalised & validated with guide ✓ Literature review complete (15-18 papers, 2021-2025) ✓ 5-7 research gaps identified ✓ Methodology designed & documented ✓ Ablation study & SOTA comparison completed ✓ Paper drafted, proofread & plagiarism-checked ✓ Target journal shortlisted — indexing & IF verified ✓ Formatted per journal template & submitted Thank You Questions & Discussion | GL Bajaj Institute of Technology & Management