Group Travel Decisions Need a Different Ranking RuleGroup Travel Decisions Need a Different Ranking Rule

The personal system uses TOPSIS, which ranks options by how close each one sits to the ideal solution. The test case was a place to buy souvenirs in Malang, with fifteen alternatives judged against three criteria.

• Distance and price act as cost criteria, so the best value is the smallest one.
• How long a shop has been standing acts as a profit criterion, so the best value is the largest one.

For a group, the team paired TOPSIS with the Borda count. Each decision maker runs their own TOPSIS pass and produces a personal order. Borda then hands the top option the largest number of points and the bottom option the smallest. The group score multiplies each alternative's closeness coefficient by its Borda points and adds the results across users.

The team compared the group recommendation against each member's personal recommendation using correlation testing.

• Three of the four users lined up closely with the group result, at 0.97, 0.99, and 0.88.
• One user sat further apart at 0.69, and the average across users came to 0.88.

The authors read that as a positive result for group recommendations built on TOPSIS and Borda, while noting the earlier personal system had reached 0.77 on its own. They suggest augmented reality as a direction for future work.