The study compared an icon-based recommender with a guide in which a language model reads the visitor and explains its choices while code keeps time and facts. Step through both versions as participants saw them, side by side.
Select a row to see the matching screens above.
| Version A · ACUX-R | Version B · Theoros |
|---|
The model chose and explained the stops; code computed every time and handled “Not for me”. Select a number to see who wrote each part.
The reason for the Currency Museum repeats an error in our place description, the archaeologist’s first name (Heinrich Schliemann). The model copied its source faithfully; the description has since been corrected.
Outside the study, Theoros also has a guide chat about the history and monuments of Athens and Attica, a camera lens for places and audio. These features were switched off during both study versions, so that the comparison covers only the profile and the recommendations. Generative features need a study link or sign-in with Google; the app works without them.
The model does two jobs only: it reads the visitor into the eight ACUX profiles, and it chooses, orders and explains stops from a list that code has already filtered. Every clock time, opening hour and walking time comes from code and verified data.
Photo likes give weights; one sentence is scored against the eight written profiles, with the evidence for each score.
Only places that can be visited in full, walk included, within opening hours and the budget. Outdoor places drop out in rain.
Ordered stops, one reason per stop from its description, one exploration stop and explained alternatives. No clock times.
Walking time and visit length give the times; stops that do not fit are dropped and gaps filled from the alternatives.
Closing, running late, rain or heat trigger one proposal at a time. No new model call.
Six planner versions on five test scenarios, two runs each. While the model wrote the clock times, adding temporal rules to the prompt made plans worse. Moving scheduling to code made all ten plans valid.
v1.0 ran at a higher thinking level with fewer checks (median latency 36 s); v1.1 to v2.2 at low thinking level, 1.9–2.6 s.
A within-subjects study: every participant did the same task with both versions, in an order the server alternates, and rated each version right after using it. Participation is anonymous; the app shows only an automatic code.
Both versions were rated highly. Version B was rated higher mainly in hedonic quality, the UEQ-S scale for stimulation and novelty, and more participants would use it on a trip. Most other ratings were at or near the top of the scale in both versions, so they could not differ much.
All participants. Hover or focus a mark for its values.
| Measure | Version A | Version B | z | p (Holm) | r |
|---|---|---|---|---|---|
| UEQ-S pragmatic (−3 to +3) | 2.50 (1.50–3.00) | 2.75 (1.75–3.00) | +1.07 | 1.000 | 0.23 |
| UEQ-S hedonic (−3 to +3) | 2.00 (1.25–3.00) | 2.25 (2.00–3.00) | +2.88 | 0.031 | 0.63 |
| Perceived accuracy (1–5) | 4 (4–5) | 4 (4–5) | +0.58 | 1.000 | 0.13 |
| Transparency (1–5) | 5 (4–5) | 5 (5–5) | +1.31 | 1.000 | 0.29 |
| Trust (1–5) | 4 (3–5) | 4 (4–5) | +0.26 | 1.000 | 0.06 |
| Confidence (1–5) | 4 (3–5) | 4 (3–5) | +0.49 | 1.000 | 0.11 |
| Use intention (1–5) | 5 (4–5) | 5 (4–5) | +0.82 | 1.000 | 0.18 |
| Profile accuracy (1–5) | 4 (3–5) | 4 (4–5) | +0.77 | 1.000 | 0.17 |
The photos definitely helped to pinpoint better my interests.Chose Version B · written in English
It makes a nice programme for my day and gives me tips.Chose Version B · translated from Greek
I like to make a general plan before my trip and have a ready list of places.Chose Version A · translated from Greek
The suggested options are quite standard.Chose Version A, about Version B · translated from Greek
The class session took place in a course taught by the researcher, who was present; the advantage of Version B was larger in that group, though not significantly. Use the buttons above to compare.