Expedia has provided you logs of customer behavior. These include what customers searched for, how they interacted with search results (click/book), whether or not the search result was a travel package.
The data in this competition is a random selection from Expedia and is not representative of the overall statistics.
Expedia is interested in predicting which hotel group a user is going to book. Expedia has in-house algorithms to form
hotel clusters
, where similar hotels for a search (based on historical price, customer star ratings, geographical locations relative to city center, etc) are grouped together. These hotel clusters serve as good identifiers to which types of hotels people are going to book, while avoiding outliers such as new hotels that don't have historical data.
Your goal of this competition is to predict the booking outcome (hotel cluster) for a user event, based on their search and other attributes associated with that user event.
The train and test datasets are split based on time: training data from 2013 and 2014, while test data are from 2015. The public/private leaderboard data are split base on time as well. Training data includes all the users in the logs, including both click events and booking events. Test data only includes booking events.
destinations.csv data consists of features extracted from hotel reviews text.
Note that some srch_destination_id's in the train/test files don't exist in the destinations.csv file. This is because some hotels are new and don't have enough features in the latent space. Your algorithm should be able to handle this missing information.
File descriptions
train.csv
- the training set
test.csv
- the test set
destinations.csv
- hotel search latent attributes
sample_submission.csv
- a sample submission file in the correct format
Data fields
train/test.csv
Column name
Description
Data type
date_time
Timestamp
string
site_name
ID of the Expedia point of sale (i.e. Expedia.com, Expedia.co.uk, Expedia.co.jp, ...)
int
posa_continent
ID of continent associated with site_name
int
user_location_country
The ID of the country the customer is located
int
user_location_region
The ID of the region the customer is located
int
user_location_city
The ID of the city the customer is located
int
orig_destination_distance
Physical distance between a hotel and a customer at the time of search. A null means the distance could not be calculated
double
user_id
ID of user
int
is_mobile
1 when a user connected from a mobile device, 0 otherwise
tinyint
is_package
1 if the click/booking was generated as a part of a package (i.e. combined with a flight), 0 otherwise
int
channel
ID of a marketing channel
int
srch_ci
Checkin date
string
srch_co
Checkout date
string
srch_adults_cnt
The number of adults specified in the hotel room
int
srch_children_cnt
The number of (extra occupancy) children specified in the hotel room
int
srch_rm_cnt
The number of hotel rooms specified in the search
int
srch_destination_id
ID of the destination where the hotel search was performed
int
srch_destination_type_id
Type of destination
int
hotel_continent
Hotel continent
int
hotel_country
Hotel country
int
hotel_market
Hotel market
int
is_booking
1 if a booking, 0 if a click
tinyint
cnt
Numer of similar events in the context of the same user session
bigint
hotel_cluster
ID of a hotel cluster
int
destinations.csv
Column name
Description
Data type
srch_destination_id
ID of the destination where the hotel search was performed
int
d1-d149
latent description of search regions
double