MODULE 1

Literature Discovery and Understanding

Discover relevant scholarly literature, explore peer-reviewed research and understand key information from academic papers using AI-assisted search, synthesis and verification.

View Full Schedule Try the Search String Builder Databases & AI Discovery Tools Publish or Perish Scopus AI Hands-On Exercise Download Module Notes (PDF)

Malaysian researchers discovering and evaluating scholarly literature in a university library

DISCOVER · EVALUATE · VERIFY

BEFORE YOU START

Come prepared for Module 1

Bring 2–3 of your own research keywords or topics to use throughout every exercise in this module.
Confirm your Scopus access (via UTM Library / EZproxy) works before the session starts.
Download and install Publish or Perish (free) from harzing.com before the session. See the download note further down this page.
Have a free Google account ready for Google Scholar searches.

No Scopus AI access? Confirm you are connected via UTM's VPN or EZproxy, or sign in again through the UTM Library portal. Remember: every AI tool in this module supports early-stage discovery, and it does not replace your own critical judgement of a study's validity.

Scan to open this module on your phone

drshahizan.github.io/bdc/module-1-literature-discovery.html

MSO

Trainer's Tip, Dr. Shahizan

Treat Scopus AI's synthesis as a map of the literature, not the destination. I always open at least the top three cited papers myself and check whether the methodology matches what the AI described, before I trust the summary in my own writing.

EXERCISE 1.1

Boolean Search String Builder ~10 min

Group your keywords into up to three concepts. The builder combines synonyms with OR and connects each concept with AND, producing a search string ready to paste into Scopus, Web of Science or Google Scholar's advanced search.

WORKED EXAMPLE

See a completed search string

Sample input: Concept 1: "student engagement, learner engagement" · Concept 2: "online learning, e-learning" · Concept 3: "higher education"

("student engagement" OR "learner engagement") AND ("online learning" OR "e-learning") AND "higher education" Search scope: Title and Abstract. Adjust field tags for your target database (e.g., TITLE-ABS-KEY(...) in Scopus, TS=(...) in Web of Science).

WHERE TO SEARCH

Start with core scholarly databases

Before turning to AI-powered discovery, run the search string you built above through the major multidisciplinary databases below. Adjust the field tags to match each platform's syntax.

MULTIDISCIPLINARY

Scopus

Comprehensive abstract and citation database covering peer-reviewed journals, books and conference proceedings.

Visit Scopus →

MULTIDISCIPLINARY

Web of Science

Clarivate citation index and discovery platform for scholarly literature, proceedings, books and datasets.

Visit Web of Science →

FREE & MULTIDISCIPLINARY

Google Scholar

Free academic search engine covering papers, theses, books, abstracts and other scholarly materials.

Visit Google Scholar →

Searching a specific field? UTM Library also provides subject-specific databases, including ACM Digital Library and IEEE Xplore for computing and engineering, ScienceDirect and SpringerLink for multidisciplinary full text, PubMed for biomedical sciences, and Emerald Insight for business and management. See the full list under AI Tools for Discovery →

BEYOND KEYWORD SEARCH

AI-powered discovery tools

These tools go beyond Boolean search. Semantic matching, citation-graph exploration and AI screening can surface relevant papers that a keyword search alone would miss. Try one alongside Scopus AI later in this module.

AI PAPER DISCOVERY

Zeta Alpha

Neural and semantic search platform for discovering and organising technical and scientific knowledge.

Visit Zeta Alpha →

AI SCHOLARLY SEARCH

Semantic Scholar

Free AI-powered search for discovering papers and following connections across scientific literature.

Visit Semantic Scholar →

CITATION DISCOVERY

Connected Papers

Builds a graph of papers related to a seed publication to reveal prior and derivative work.

Visit Connected Papers →

AI LITERATURE SEARCH

Elicit

Finds papers, supports study screening and extracts information into structured review tables.

Visit Elicit →

EXERCISE 1.2

Rank papers by citation impact with Publish or Perish ~10 min

Publish or Perish is a free desktop application that retrieves citations from Google Scholar, Scopus, Crossref and other sources, then calculates bibliometric indicators such as total citations, h-index and citations per year, helping you spot the most influential papers and authors before you commit to reading them.

Search

Paste the search string you built above into Publish or Perish, or search directly by author, topic or journal.

Rank

Sort results by citations per year to surface the most influential and most recent papers first.

Export

Export the reference list to BibTeX, RIS or CSV for your reference manager.

Free download for this exercise: Install Publish or Perish from harzing.com/resources/publish-or-perish before the session. A citation count is one signal of impact, not a substitute for reading the paper.

EXERCISE 1.3

Try Scopus AI with your own topic ~15 min

Scopus AI lets you ask a plain-language research question and get a synthesised answer drawn from Scopus-indexed, peer-reviewed literature, along with suggested references and follow-up questions. Use this exercise during the session with your own research topic.

Access

Log in to Scopus AI through your institution's Scopus subscription (available via the UTM Library) or your personal Scopus account with AI features enabled.

Ask

Type one natural-language question about your own research topic. Avoid Boolean operators and keyword strings, and ask as you would ask a colleague.

Verify

Open the papers Scopus AI cites and check the abstract and methodology before treating any claim as established.

TOPIC + OBJECTIVE + CONTEXT

Universal question pattern

What are the [RESEARCH OBJECTIVE] regarding [TOPIC] in [CONTEXT OR POPULATION] over [OPTIONAL TIME BOUNDARY]?

Example: What are the emerging applications of explainable artificial intelligence for medical image diagnosis over the last five years?

DURING THIS SESSION

Hands-on checklist

Ask Scopus AI your main research question using the pattern above, with your own topic.
Note down 3 papers Scopus AI surfaces that look genuinely relevant.
Ask one follow-up question to explore a gap, trend or contradiction.
Open one suggested paper and confirm the abstract actually supports what Scopus AI said.

Verify the evidence: Scopus AI supports literature discovery, not final scholarly judgement. Open the original publications and evaluate their design, sample, context, limitations and full-text findings before using or citing a claim.

For the complete library of 48 Scopus AI prompts across 12 categories, covering gaps, trends, methods and article positioning, see Scopus AI Prompts for Research Discovery →

EXERCISE 1.4

Citation snowballing ~15 min

Snowballing finds papers a keyword search misses, by following the citation trail of one strong paper instead of guessing more search terms.

Pick one strong seed paper from your search results, ideally one that is highly relevant and reasonably well cited.
Snowball forward: on Google Scholar, open the seed paper and click "Cited by" to see newer papers that cite it. Note 2 to 3 that look relevant.
Snowball backward: open the seed paper's own reference list and note 2 to 3 references that look relevant to your topic.
Record each new paper below, then repeat one more round from your best new find if time allows.

Round	Direction	Paper (author, year)	Why it is relevant
1	Forward (cited by)	e.g., Wong 2023	Extends the seed paper's method to a new population
1	Backward (reference list)	e.g., Tan 2019	Original source of the framework the seed paper uses

Reveal a worked example

Seed paper: "Explainable AI for Medical Image Diagnosis: A Systematic Review" (Lim et al., 2022).

Forward (cited by): A 2024 paper applying the same explainability framework to radiology datasets. Relevant because it shows how the seed paper's method has since been extended.
Backward (reference list): The earlier paper that first proposed the explainability technique the seed paper builds on. Relevant because it is the foundational source the seed paper depends on.
Common mistake to avoid: Recording every citing or cited paper without checking relevance first. Snowballing should still be selective, skim the title and abstract before adding a paper to your table.

EXERCISE 1.5

Compare database coverage ~15 min

No single database indexes everything. Running the same search in more than one database shows you what you would have missed by relying on just one.

Take the search string you built above and run it, adjusting field tags as needed, in two or three databases (e.g., Scopus, Web of Science, Google Scholar, or a subject database from your field).
Record the total result count and the top result in each database.
Check whether the top results overlap across databases, or whether each surfaces different papers.
Note one likely reason for any difference (e.g., subject coverage, indexing lag, grey literature included or excluded).

Database	Total results	Top result	Notable difference
e.g., Scopus	e.g., 214	e.g., Smith 2021	e.g., Strong in engineering & CS
e.g., Google Scholar	e.g., 1,860	e.g., Lee 2022	e.g., Includes theses & preprints

EXERCISE 1.6

Check a source with the CRAAP test ~10 min

Before you rely on a source, especially one an AI tool suggested, run it through the CRAAP test: Currency, Relevance, Authority, Accuracy, Purpose.

Currency: When was it published or last updated? Is that recent enough for your topic?
Relevance: Does it actually address your research question, or just share a keyword?
Authority: Who are the authors, and what is their expertise or affiliation? Is the venue peer-reviewed?
Accuracy: Is it well-referenced, and can you verify its claims or data against other sources?
Purpose: Why was it written? Is there a bias, funding source or agenda you should account for?

SOURCE CHECK PROMPT

Ready-to-use prompt

Act as a research methods advisor helping me evaluate a source.

Source details (title, authors, year, venue, and a short summary or abstract):
[PASTE SOURCE DETAILS HERE]

Assess this source using the CRAAP test: Currency, Relevance, Authority, Accuracy and Purpose. For each criterion, give a short assessment based only on the details I provided, and flag anything you cannot assess from the information given rather than guessing.

End with an overall recommendation: safe to cite, cite with caution, or verify further before relying on it.

RELATED TOOLS & PROMPTS

Continue your literature discovery

DISCOVERY & DATABASES

AI Tools for Discovery

Curated academic search engines, citation databases and literature-mapping tools.

Explore Discovery Tools → 48 FOCUSED QUESTIONS

Scopus AI Prompts for Research Discovery

Copy-ready questions to explore trends, gaps, methods and article positioning.

Explore Scopus AI Prompts → 95 SOURCE-GROUNDED PROMPTS

Gemini Notebook Prompts

Analyse and compare uploaded papers once you have shortlisted your sources.

Explore Gemini Notebook Prompts →

WHY USE AI HERE

Traditional vs AI-assisted literature discovery

Task	Traditional approach	AI-assisted approach
Building a search string	Trial-and-error keyword combinations, one database at a time	Search String Builder generates a complete Boolean string instantly
Finding related papers	Follow each reference list manually, one citation at a time	Connected Papers or Zeta Alpha map related work instantly as a graph
Judging a paper's impact	Look up citation counts one paper at a time across separate sites	Publish or Perish calculates h-index and citations per year in one pass
Understanding findings across papers	Read every abstract individually and synthesise by hand	Scopus AI synthesises an answer across dozens of papers at once

The trade-off: AI tools save time on breadth; you still verify depth. Always open the underlying papers before treating a synthesised claim as established.

RECAP

Key takeaways and terms

Key Takeaways

A good Boolean search string groups synonyms with OR and connects concepts with AND.
Start with core multidisciplinary databases, then widen with AI-powered discovery tools.
Citation metrics (Publish or Perish) rank impact, but never replace reading the paper.
Always verify a Scopus AI synthesis against the original, cited publications.

Boolean operators: The operators AND, OR and NOT, used to combine or exclude search terms.
h-index: A metric measuring both productivity and citation impact of an author's publications.
Citations per year: Total citations divided by years since publication, which normalises impact across paper age.
Semantic search: Search based on meaning and context rather than exact keyword matches.

WHAT'S NEXT

Have a shortlisted topic and a handful of strong papers? Module 2 helps you turn that into a clear, FINER-checked research question.

5 MODULES · 1 DAY

See how Module 1 fits into the full programme.

Review timings for every module, breaks and the closing wrap-up.

View Full Schedule