Using Social Wisdom to help travelers select their next destination
COVID-19 is winding down and traveling is back with travelers in the pursuit of a trip of a lifetime. Can machine learning help travelers on their journey to transformative and meaningful travel experiences?
As we enter the long tail of the COVID 19 pandemic we are witnessing records in the number of travelers looking to travel. According to Expedia — 2022 will be the year of the GOAT (Greatest of all Trips) in the pursuit of transformative and meaningful travel experiences. Whatever the motivation travelers can gain access to detailed experiences of other travelers in the form of user-generated reviews opinions photographs, and videos all contributed through online travel platforms. The wealth of available data is confounding rich with bias and usually not provided by an honest broker. In the following article, I will review the challenges and opportunities for using machine learning to aid travelers to plan the trip of their lifetime.
A fragmented Traveler Journey
Massive travel reviews of tourist sites are becoming easily accessible through social networks, such as Yelp, Google, TripAdvisor, foursquare, and so on. These reviews support the different types of information about visited attractions, visited times, travel notes and basic profiles of travelers, labels, ranks, review texts, and basic attributes of attractions. In principle, these sites empower the travel community allowing users to bypass financially-incentivized travel agents, activity agents, and tourism bureaus. Over time as the amount of travel information online and the number of travel tools has grown, so has the complexity of planning and booking a trip. The perceived value of these reviews is reflected in the millions of individuals producing and consuming content related to travel destinations and tourist activities. Nielsen research found that travelers spent an average of 53 days visiting 28 different websites throughout 76 online sessions, with more than 50% of travelers checking social media for travel tips. Discovering things to do or altering plans during a trip further adds to this complexity.
Information overload
The huge amount of reviews from experienced travelers has a profound impact on the choice of destinations by travelers. However, the increase in global tourism (putting COVID aside) with the increase in the number of travelers has led to a rapid increase in the number of online reviews. It is difficult and even impossible for potential travelers to browse and analyze attractions reviews in detail. A large amount of redundant and meaningless information interferes with the judgment of potential travelers creating a massive information overload problem. Not to mention, fake data that appears on many of these travel reviews sites as was the case with Tripadvisor which warrants not relying on one single source.
When a potential traveler has a desire to travel somewhere, he or she has a set of possible destinations in mind. However, the potential traveler is hesitant to choose between several alternatives as he has limited knowledge of the alternatives and does not necessarily possess the expertise required to prioritize and choose a destination. Hence he starts reading online reviews about the alternatives. The potential traveler wants to make the most comprehensive assessment and analysis of alternative destinations from various aspects but usually ignores some significant aspects due to the limited understanding, time constraints, or laziness which may result in decision mistakes, i.e not choosing the most favorable location.
Using Social Wisdom to help a bewildered traveler
To solve the travel decision-making problem of potential travelers, we need to provide travelers with a decision-making aide for ranking attractions and reduce the risk of decision mistakes. To this end, we can extract meaningful and useful information from the tremendous amount of online reviews and then rank alternatives taking into account the traveler’s individual preferences.
To build such an effective decision support system we need to do the following:
- Create a web crawler to crawl and preprocess online reviews.
- Extract the aspects that concern the traveler and calculate the sentiment orientation.
- Determining the weights and criteria for the ranking system.
The main challenge with this flow is analyzing the traveler reviews. The sentence structure is complex lacks organization and suffers from bias which makes it difficult to solve using traditional methods.
Sentiment analysis has been a new analytical method in NLP in recent years. This method is mainly used for analyzing texts with emotional contents and for judging sentiment polarity. Sentiment analysis is divided into three levels: document-level, sentence-level, and aspect-level. The document level identifies the sentiment polarity of the whole document, The sentence level identifies the sentiment polarity in each sentence of the text. Both of the above two levels of sentiment analysis are devoted to the polarity classification of opinions, but ignore the extraction of evaluation objects or targets. Aspect-level sentiment analysis (i.e., fine-grained level) identifies the sentiment polarity corresponding to each aspect of the text. In a review, travelers may offer their opinions on food, transportation, culture, and other aspects of a destination. The extracted factors may be positive, neutral, negative, or other categories. All well-known artificial intelligence (AI) service platforms provide services to enable similar analyses. Current orientation analysis has limitations as it tends to provide a score that doesn’t necessarily portray the decisiveness or ambiguity of these reviews (but this is a discussion for another article).
The ranking will be determined based on the weighted traveler preferences and review score. Different reviews have different levels of importance based on factors such as date of review, helpfulness votes as well, the total number of reviews, and reviewer experience (based on the number of posts).
Building Trust
Once we pass the technical barriers to creating an effective attraction Social Intelligence ranking system (which is much more complex than the short description provided) we need to gain the traveler’s trust. If the whole point is to help the traveler make a more informed decision based on the unlimited data available how will we gain his trust?
This touches on a much broader point on how to build collaboration between humans and AI but for the sake of our discussions, I think we need to offer the travelers ways to input (collaborate) with the ranking system through an onboarding process where the traveler provides his personal preferences and can sense how this impacts the destination and attractions ranking. Another key aspect is the ability to share and get feedback from friends and the expert community on the ranking result. This will provide more feedback to evaluate the level of the recommendation system (feedback that will further improve it)
What’s next
Looking forward to the sheer amount of data being produced offer an unparalleled opportunity to better understand the traveler’s experience. This will help travelers create personalized itineraries that will help them avoid the endless online search, fragmented experiences, travel anxiety, and improve their decision-making skills. The large amount of data available, a variety of users across geographies, interests, and experience levels provide the basis for additional insights on attractions and matching these to the right user. These insights into the world of travelers are critical for the travel industry, local government, and related Destination Marketing Organizations. They will also allow for better targeting of advertising campaigns.
Sources
https://link.springer.com/article/10.1007/s40815-021-01131-9
https://www.granthaalayahpublication.org/journals/index.php/granthaalayah/article/view/4178/4266
https://jespublication.com/upload/2022-V13I3032.pdf
https://ieeexplore.ieee.org/document/9102673
https://newsroom.expedia.com/2021-11-30-The-GOAT-mindset-Expedia-reveals-2022s-biggest-travel-trend
https://www.electrifai.net/blog/applied-ai-helps-companies-with-post-covid-travel-rush
https://business.adobe.com/au/resources/digital-trends-in-travel-and-hospitality.html
