How to read AKTU PYQs like data, not like a PDF

Most students “do PYQs” by opening five years of papers and reading them top to bottom. By paper three, every question starts to look important, and you close the PDFs with the same feeling you started with: everything might come.

Previous-year questions are far more useful when you treat them as data. You do not need software for this — a notebook or a spreadsheet is enough. This post walks through the method we use to build PaperMap’s heatmaps, so you can do it by hand for any subject.

Why reading papers top to bottom doesn’t work

A question paper is organised by section and question number, not by topic. That is the wrong shape for revision. When you read papers in that order:

• You remember the last paper you read, not the most frequent topics.
• Reworded repeats look like different questions, so a topic asked four times feels like four separate things to learn.
• Topics that were never asked are invisible — you cannot see an absence while reading.

Flipping the shape from paper → questions to topic → times asked fixes all three.

Step 1: Write the syllabus down as topics

Take the official syllabus for your subject from aktu.ac.in and break each unit into 4–8 topics. Keep them small enough that one topic fits one answer. For BCS303, “Unit 2: Algebraic structures” is too big; “Groups and subgroups”, “Lagrange’s theorem”, “Rings and fields” are the right size.

Step 2: Tag every question with a topic

Go through each paper once and write, for every question: the year, the section, the marks and the topic it belongs to. That’s it. Do not solve anything yet.

A simple sheet looks like this:

year,section,question,marks,unit,topic
2023-24,A,1(c),2,2,Groups and subgroups
2023-24,B,2(b),7,5,Recurrence relations
2022-23,C,4(a),7,3,Lattices
2022-23,A,1(f),2,2,Groups and subgroups

Step 3: Merge reworded repeats

Examiners rarely copy a question word for word. They rephrase it. Before you count anything, group questions that ask for the same answer:

BCS401Unit 3· 7 marksexample

Explain deadlock prevention techniques.

The three wordings above need one prepared answer. When you count, they are one topic asked three times, not three topics.

Step 4: Count, then sort

Now build a second, tiny table with one row per topic:

Topic	Times asked	Last asked	Total marks
Groups and subgroups	4	2023-24	18
Recurrence relations	3	2023-24	21
Rings and fields	0	—	0

(Example numbers to show the format — fill in your own from real papers.)

Sort by total marks first, then by how recently the topic was asked. The top third of this list is where your first study hours should go.

Step 5: Look at what’s missing

The most underrated row in the table is the one with zero. A topic that has not been asked in five years is not guaranteed to stay absent — but if you have six days left and twelve topics to cover, it is a reasonable one to move to the end.

What this gives you

After about two hours per subject you have:

1. A ranked topic list backed by real papers, not someone’s opinion.
2. One question per topic with all its wordings, ready to prepare a single strong answer.
3. A short list of low-yield topics you can consciously push to the end.

That is exactly what PaperMap’s Topic Heatmap and Repeat Detector automate — for every subject, with the mapping already done. But the method works with a pen, and it is worth knowing how the numbers are made, whichever tool you use.