AI-Powered Criminal Network Analysis System

The receipts 🧾

Every red flag on this statement, in full. These are the four places it bites.

1. Exhibit A

   No real investigation dataset exists, so you synthesise FIRs and call records and the network only reveals the connections you built in

2. It gets worse

   Commercial link-analysis tools already do this, so the novelty bar for a student build is high

3. Still reading?

   Entity extraction from messy real police reports is the actual bottleneck and cannot be shown on clean synthetic text

4. And the finisher

   Automatically flagging individuals as key criminal actors carries real consequences if the extraction or linking is wrong

The damage report 📉

Every score this statement earned, and what each one actually costs you.

• Feasibility

  3/5

  Buildable. Not comfortably. There is a week in here you have not planned for yet.

  Entity extraction, graph construction and centrality analysis are standard, but there is no real criminal-investigation dataset so you synthesise FIRs and call records, and the value of the network analysis depends heavily on how realistic and interconnected that synthetic data is.

• Innovation scope

  3/5

  Mildly interesting. The novelty will not carry the room; the build has to.

  Entity extraction plus graph analytics is an established pattern, so your room is in the multi-source fusion and the suspicious-pattern detection rather than in the concept, which the description largely specifies.

• Clarity

  4/5

  The ask is unambiguous, which quietly removes your favourite excuse.

  The data sources, the entities to extract, the relationship-mapping, key-individual and pattern-detection requirements and the visual output are enumerated clearly, so the target is well defined even without data.

• Acceptance potential

  3/5

  Middle of the pack. This statement will not win the room for you — you will have to.

  The link-analysis concept is genuinely valuable and demos well, but tools like this already exist commercially, you must synthesise the investigation data so the network only reveals what you built in, and the entity-extraction accuracy on messy real reports — the actual bottleneck — cannot be shown on clean synthetic data.

• Effort

  Massive

  A semester of work wearing a hackathon costume. Something is getting cut; decide what now, not in week five.

  Multi-source ingestion, NLP entity extraction, graph construction, centrality and pattern detection and an interactive visualisation is a broad, multi-part build.

• Demo-ability

  Easy

  Easy to demo — and so is everyone else's. Working is the floor here, not the achievement.

  A relationship graph revealing the central individual linking separate clusters is a compelling, investigative-looking demo that reads instantly.

• Data

  None supplied

  No dataset comes with this one, so every accuracy figure you quote is a number about labels you invented.

  Nothing is provided with the statement. You are sourcing, cleaning and labelling it yourself, and that work is invisible in the demo but very visible in the questions.

The demo they will have already seen 🪞

Somewhere around 70–160 teams are heading here, and the description is doing the choosing for most of them. They will read the same brief, reach the same architecture, and build a version of the same demo you are planning. Being correct is the floor. If your five minutes could be swapped with the team before you and nobody in the room would notice, you have not picked badly — you have built predictably, which costs exactly the same and hurts more.

What survives 🛡️

The ground worth standing on when the questions start.

• Entity extraction and graph analytics are mature with strong libraries, so the core is buildable and reliable
• A graph revealing the central connector between clusters is a compelling, self-explanatory investigative demo
• Multi-source fusion linking phone, financial and report data is genuinely useful and the real value beyond single-source analysis