"Big Data"--for the tourism industry can be unlimited

Reading: The ability of large data technology to discover patterns in large amounts of data clearly offers a lot of opportunities for tourism, where companies with large data projects can sell tourism products to consumers in a new way.

If you are a loyal reader of Tnooz, you must have read about the potential of big data.

But like most emerging technologies, people are confused about what it really means, can it bring a better customer experience and boost sales?

This article is just a tip, analyzing some potential applications of large data in the future of tourism:

Large data

The word "big data" is usually expressed in several ways, and its meaning is self-evident. "Big data" means a lot of data, that is, in terms of scale, workload and overall cost, it goes beyond the data set of the common database, and the technology that extracts the meaning of the analysis.

So where do we meet a lot of data in the tourism industry?

One of the best examples is the online travel agency's Analytics log. Over the years, analysis tools have allowed travel companies to track conversion channels, detailed consumer data, and other relevant information, such as which pages have the highest conversion rates? Which pages have the highest bounce rate and so on.

These analysis reports are then used to optimize the site to ensure the highest conversion rate. But in the era of big data, tourism companies are collecting and focusing more data than ever before.

For example, some websites also collect real-time information about the specific mouse page activity of visitors.

For each user naturally generated a lot of information, so that tourism enterprises for each user on the page behavior has a unique understanding.

A few years ago, it was even harder to complete such data storage (because it was very expensive).

Today, cheap storage is becoming popular, and the Distributed file system allows data to be shared between dozens of and even hundreds of of computers, thus enabling cheap and efficient data storage.

As storage technology improves, the cost of storing each byte of data continues to decline.

Large data analysis, mapreduce and semantic extraction

It's good to have all the data, but the real value depends on the semantic extraction feature. Based on the technology originally invented by Google, large data tools such as MapReduce can easily discover common trends in the data, as well as behavior corresponding to these trends.

For example, it's easier to understand:

Imagine that you have an Excel form that contains all the hotels in the world, and you want to find out which ones are described as "great!" "The hotel.

The original data may look like the following figure:

Hotel Name--description

—————————–

Hotel a "bad experience"

Hotel B "Great pool"

C Hotel "like"

Hotel B "Great Restaurant"

A hotel "Love"

Hotel C "great experience"

B Hotel "boring"

Here are just a few of the records, which may have hundreds of or even thousands of pages.

With MapReduce, you can write a feature that identifies each hotel name that meets the criteria and integrates a set of comments for each hotel.

In the example above, "Hotel B" appeared 3 times, so the map feature might create the following set of information:

B Hotel: ' Great swimming pool ' great restaurant ' boring '

So the map feature helps us find all the reviews for Hotel B, but that's not enough, and we need to use the reduce feature for the next analysis.

This work can help us perform various analyses based on the dataset created by map. In this case, we want to find a comment that contains only the word "great". The reduce feature writes the following computer code:

"If you find that the comment contains a great word, the internal counter is 1." ”

Internal counters usually show scores, or the number of "great" words in a particular dataset. In this case, the B Hotel will be ranked first (two "great"), then the C hotel, and finally a hotel.

So what do we get in this case? We found common ground based on raw unstructured data and analyzed it for business purposes.

Although seemingly simple, this is the power of MapReduce and related technologies: extracting the rules that are hidden in the data.

The amazing thing about MapReduce is that it can analyze the data in hundreds of billions of dollars and then discover the rules.

Can you feel this potential now?

The state of large data in the field of tourism

The ability of large data technology to discover rules in large amounts of data clearly provides a lot of opportunities for tourism to reshape marketing and sales practices for consumers.

There is no doubt that companies with large data projects can sell tourism products to consumers in a new way.

Emmanuel Marchal, currently in Acunu, a large data storage company, briefly describes the state of tourism's large data:

Large data applications are transitioning from the analysis phase to the true personalization phase. For example, true personalization means that a travel site can offer different hotel choices to different travelers based on their specific needs, preferences, and previous purchases, rather than providing a popular choice on the basis of their categories.

The real individuality is the main driving force, is the tourism big data "the most important".

OTA and other tourism companies should regard large data as "necessary" rather than "dispensable".

On large data, 2009 began to talk about the 2010, many companies began to start large data projects, the prototype in 2011, then 2012 Annual meeting is the beginning of extensive practice of large data projects? In addition to personalization, Hopper has tried this and provided another example of how big data serves consumers.

Large-scale analysis of the combination of natural language processing and location makes it possible to search for "seaside resorts under 500 dollars nearby". While this requires technology consolidation, large data-processing applications make them possible.

Looking to the future

Although personalization is "the most important", there are a lot of potential users who use large data in tourism.

Geo-fencing (please click here to comment) technology appears, which can be used to monitor travellers near a particular attraction or merchant.

The recently introduced Foursquare radar feature is a good example of this application, and when you want to go somewhere and set it up, your phone automatically alerts you when you're near it.

The technology is big data: Search for real-time geographic location information through your mobile's GPS function, and then remind you to be in a specific area.

This feature is available to hundreds of millions of consumers, and the amount of data that needs to be collected and processed during this process was unimaginable a few years ago.

Imagine how many users use the GPS navigation function every day? The potential of geo-fencing technology for marketers is enormous.

Image data processing technology: Every day in billions of photos uploaded on the Internet, which represents a huge capacity.

Picture-sharing applications such as Instagram are a great way to share wonderful moments with friends, but the real value of these startups is the data they collect in the background. Each image contains a large amount of data that can be further analyzed to identify the user's main information.

Color, the controversial venture has been priced at sky-high prices before its launch, and the huge potential of the data they want to collect has persuaded investors.

In the tourism industry, the Enterprise Jetpac has already adopted the image processing technology, has injected the different color to the social tourism domain.

Hunch, a start-up enterprise recently bought by ebay, is also a good example of finding patterns in scattered data.

Their APIs and some of the public testing tools let everyone find a similar "40% like Subaru car people like to stay in five-star hotels" conclusion. (This is not an example, of course)

While this correlation may seem suspicious, imagine how valuable this correlation would be if you were trying to cross-sell a traveler.

ParknShop graphs will understand the customer's "something" and become aware of "many other things" at the same time. The use of ParknShop graphs applications will certainly be widely used in the future, from online retailers to tourism.

In short, big data is not an amazing magic box that you can buy from a vendor.

Large data is the concept that we generate a lot of data every day, and with the development of technology applications, we can better sell products and services.

Are you ready?

(Responsible editor: Lu Guang)