4

I'm building the search page for an OTA (online travel agency). The brief has a few quirks that make this task difficult, so I need some help to identify some feasible ideas.

Most OTAs have a "standard" interface with a form asking for dates, places, and the number of rooms or travelers. Then they return a list on the right with a large number of filters available on the left.

This has a number of disadvantages:

  • one needs to know a number of exact pieces of data in order to search
  • filters are complex and highly useful criteria go mixed with marginally useful ones
  • there is almost no space for fuzzy, imprecise criteria such as "I want to travel one of the next weekends" or "I want to go to Europe with my girlfriend for less than 500 Euro"
  • it's almost impossible to book trips, which have multiple parts (e.g. hotels, transports, airplane, activities...) unless they are pre-made packages on offer.

What I need to build needs to:

  • allow for fuzzy and incomplete criteria
  • allow for complex queries involving different flights for different people, for example, total budget, and other complex requirements
  • not to be a fixed form like other OTAs
  • not to be a textbox where one writes their wishes and magically a result appears.

What is normally considered best practice to allow the gathering of fuzzy, imprecise, or incomplete requirements from a user in a website or mobile?

Improve this question
1
  • 4
    I recommend you pare down your question. As it stands, it's reads to me like, "What is the best practice for building a site that's better than Expedia, and with only one search box?" Are you asking about which technology to employ to start with a sentence and provide the user travel recommendations? That's incredibly complex, but it's a more answerable question (which could be NLP, Google it ;) When you figure out how to get to Europe with your girlfriend for less than 500 €, let me know!
    – Joel Mellon
    Oct 3, 2021 at 15:06

1 Answer 1

Reset to default
1

I think 'fuzzy' search is a term that was used when natural language processing wasn't as mature as it is now. But I suspect that you are looking at having to come up with a simpler problem to solve, or to spend the time and effort in creating a more sophisticated solution. It all comes down to time and $...

I suggest looking at Wolfram Alpha and their examples of Natural Language Processing capabilities as to how they manage the expressions that can be used as input and how it is processed.

There are probably other examples as well but I think this is an area that they seem to specialise in and has been successful in creating proper context for different types of searches.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.