Telia Crowd Insights

Activity data 

Activity data reports the number of people dwelling in area for a certain amount of time. More precisely, activity count calculates the number of unique persons with an activity per spatial and temporal unit. 

Here, the spatial unit is one Telia scalable grid cell, and time unit is a day. An activity is identified if a person is dwelling at least 20 minutes in the same area. Therefore, counts refer to distinct persons only within a time slot and grid cell. 

The home zone is based on the longest activity before 09:00 for each person (daily). Should a person not have a dwell, it is based on the first ping(s). The spatial unit for a home zone is a postal code in Finland. One person can have one home zone per day. Due to Telia’s privacy policy all daily activity home zone- combinations with less than 5 people are removed from the dataset. 

Data set: Aalto_activity_homezone.csv 

• Time resolution: Daily. 
• Area: Activity in the capital region of Helsinki (Helsinki, Vantaa, Espoo and Kauniainen). 
  • Home zone information in Finland on postal code level (Statistics Finland 2019). 

Data format

• date: date when the activity happened. 
• people: number of people with an activity. 
  • These numbers are extrapolated to represent the whole population, based on the daily observed market share (Telia) and population statistics (Statistics Finland). 
• grid_code: grid id where the activity has taken place. 
• grid_name: grid name (based on nearby locations). 
• grid_size: size of the grid in meters (500 = 500 x 500-meter grid). 
• home_al3_code: Home zone code on postal code level. 
• home_al3_name: home zone name on postal code level. 
• home_al2_code: home zone code on municipality level. 
• home_al2_name: home zone name on municipality level. 
• home_al1_code: home zone code on region level. 
• home_al1_name: home zone name on region level. 
• fin_daily_rating: a daily rating based on the signal quality of the day. 
  • The rating ranges from A (best) to D (worst). The classification is based on a signal drop percentage from the baseline: A: >= -5%, B: => -10% and C: >= -15%. 
  • D-rated days < -15% are removed by default from this dataset, as well as dates that are missing."