Arjen aktiviteettimieltymykset pääkapunkiseudulla

Interest in examining people in an urban environment through the places they visit during their everyday life has increased. While earlier studies have focused more on people’s residential locations, the approaches considering the whole activity space, capture urban behaviour more thoroughly. The shift to this kind of analysis is enabled by the increased availability of location data from mobile phones and other devices. In this thesis, mobile application usage data is utilised to form an overall view of people’s daily habits in the Helsinki Metropolitan Area, with a focus on the months August to October in 2024. The thesis aims especially to detect possible spatial patterns in activity preferences and to analyse how activity patterns are linked to socioeconomic characteristics. The dataset used describes individuals’ likelihood of visiting different categories of points of interest (POI), such as shopping or swimming pools. The likelihood is presented in the dataset as an affinity score. Since the dataset was not customised to the Helsinki Metropolitan Area or the aim of this study, multiple preprocessing steps were completed to adapt it to the scope. These included, for example, filtering users and enriching the data with socioeconomic indicators for postal code areas. The results of the thesis suggest that there are detectable spatial and socioeconomic patterns in activity preferences. The preferences varied, for example, depending on the neighbourhood’s centrality, income level or level of family-orientation. Shopping and travel activities were more popular in central neighbourhoods, while the popularity of education, health and leisure activities was more evenly distributed over the study area. Higher income, family-oriented, and larger-dwelling neighbourhoods tended to have lower visitation values across all activity categories. In contrast, younger, adult-dominated neighbourhoods were associated with a higher likelihood of visitation to different categories. The data also included some limitations, such as possible demographic bias in mobile application users and insufficient coverage of the study area at the neighbourhood level. In addition, users’ home or work locations can possibly distort observed activity preferences.