Uncovering Tourism in Jordan through machine learning and radio mining

By Ayah Younis, Head of Exploration, UNDP Jordan Accelerator Lab

February 16, 2023

Since June 2022, the UNDP Jordan Accelerator Lab has been exploring, mapping, and experimenting with knowledge and solutions that could improve tourism in Madaba while creating income for the local community. We collected qualitative data typically unavailable to decision-makers. In this series, we will share the learning from these activities.  

As a traveller, have you ever scrolled through TripAdvisor reviews to get an idea of what to expect from your travel destination? Have you ever wondered how tourism trends are formed from these online reviews? The Jordan Accelerator Lab is doing just that, and we are using machine learning!

The TripAdvisor sentiment analysis activity aims to gather online data from TripAdvisor to generate tourism trends in Jordan, specifically in Madaba. The project utilised machine learning to scrape text from forum posts and comments on the website and used artificial intelligence to conduct a sentiment analysis of these posts based on predefined attributes. The machine learning system categorised the text by attributes and identified the sentiment related to each aspect, such as transportation, food, service, and more.

The results of this project were astounding. The machine collected over 70,000 entries in English about tourism in Jordan, from which 20,000 entries were relevant to the set attributes. The data collected was then grouped into an interactive dashboard, which provides insight into the travellers' experiences in Jordan. The volume of data collected through this method is beyond the ability of manual data collection and analysis.

Through machine learning and sentiment analysis, the Jordan Accelerator Lab is shedding light on the tourism trends in Jordan. This project provides insight into what travellers love about Jordan and what they'd like to see improved. This information can be used to enhance tourism experiences and make travelling to Jordan an even more enjoyable experience.

The second activity undertaken by the Jordan Accelerator Lab is Radio mining. In this project, the Lab bought airtime from Mazaj FM, a private local radio station. During a 20-minute segment on the daily talk show "Al Taree2" with Ma'en and Haneen, the presenters discussed Madaba as the Arab City for Tourism for 2022. Listeners were encouraged to answer questions the Lab had provided on local tourism. 



The learnings from radio mining were equally valuable. There is insufficient media coverage or promotional campaigns about Jordanian cities that are acknowledged regionally or globally. The presenters stressed the importance of raising the local population's awareness of these cities' historical and cultural significance, which would positively impact local engagement with tourism activities. To tackle this aspect, some participants recommended that school education is one element that needs to be considered.

The presenters also highlighted the importance of digital marketing on social media as an excellent tool to market local tourist experiences. However, the participants shared their concerns about the high costs of local tourism and their preferences to travel abroad with the same budget they'd spend in Jordan. 

The Jordan Accelerator Lab's TripAdvisor sentiment analysis and radio mining projects were valuable and unconventional tools for understanding the tourism trends in Jordan. Through machine learning and radio, the Lab is collecting data and providing valuable insights that can be used to improve tourism experiences in the country. By better understanding the tourists' preferences and behaviours, Jordan can better cater to their needs and enhance their overall travel experience.