possible dependent variable: [reach_per_million, page_views_per_million, rank]
preferred variable: reach_per_million [the rank is global and all the available data have a low rank]
- Reach per million value - This metric means "out of a million global internet users, how many visited this site (during the specified time range)?" replaced by views
- Page views per million value - This metric means "out of a million pageviews made by the Alexa traffic panel (during the specified time range), how many were on this site?" Note that the pageviews are not unique.
- hashtags: how many hashtags did this account include
- likes: how many likes did this account recieve
- links: how many links to their website did this account include
- replies: how many replies did this account recieve
- retweets: how many retweets did this account recieve
- total_tweets: how many tweets did this account tweet
- word_count: count of words *is_weekend: if this date is weekend or not
- links_in_count: A count of links pointing in to this site (no historic data available only current value)
- followers_count: total followers count
Summary of attributes We have 6956 observations and 52 features including dummy variables
| Column Name | type | Description |
|---|---|---|
| views | float | Number of visitors (A visit is defined as an entry to a web domain from a different web domain or from the beginning of an empty browsing session, and expires after 30 minutes of inactivity) |
| date | date | The date of the traffic data\twitter |
| category | string | type of website (News, Shopping etc... ) |
| location_country_code | string | two digit country code of site origin |
| twitter_data | numbers | (followers_count, hashtags_count, likes, links, mentions_count, replies, retweets, total_tweets ) |
| links_in_count | int | The number of links for this site found on sites |
- Dealing with multicollinearity using variance inflation factor (VIF) threshold of 5
- Transforming the target variable using log1p
- Creating dummy variable for location + category
- Generate is_weekend attribute from date
- Forward and backward fill for missing values in our dependant variable