A spatiotemporal knowledge graph-based method for identifying individual activity locations from mobile phone data

Jian Li,Tian Gan,Weifeng Li,Yuhang Liu

2024-10-17

Abstract:In recent years, mobile phone data has been widely used for human mobility analytics. Identifying individual activity locations is the fundamental step for mobile phone data processing. Current methods typically aggregate spatially adjacent location records over multiple days to identify activity locations. However, only considering spatial relationships while overlooking temporal ones may lead to inaccurate activity location identification, and also affect activity pattern analysis. In this study, we propose a spatiotemporal knowledge graph-based (STKG) method for identifying activity locations from mobile phone data. An STKG is designed and constructed to describe individual mobility characteristics. The spatial and temporal relationships of individual stays are inferred and transformed into a spatiotemporal graph. The modularity-optimization community detection algorithm is applied to identify stays with dense spatiotemporal relationships, which are considering as activity locations. A case study in Shanghai was conducted to verify the performance of the proposed method. The results show that compared with two baseline methods, the STKG-based method can limit an additional 45% of activity locations with the longest daytime stay within a reasonable spatial range; In addition, the STKG-based method exhibit lower variance in the start and end times of activities across different days, performing approximately 10% to 20% better than the two baseline methods. Moreover, the STKG-based method effectively distinguishes between locations that are geographically close but exhibit different temporal patterns. These findings demonstrate the effectiveness of STKG-based method in enhancing both spatial precision and temporal consistency.

Social and Information Networks,Physics and Society

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to solve the problem of insufficient consideration of spatial and temporal relationships when identifying individual activity locations from mobile phone data. Specifically, existing methods usually aggregate location records in multiple days only based on spatial correlation to identify activity locations. However, this method ignores temporal relationships, which may lead to inaccurate identification of activity locations and affect activity pattern analysis.

#### Specific problems include:
1. **Only considering spatial relationships while ignoring temporal relationships**: Current methods mainly rely on spatial proximity to identify activity locations, but ignore temporal associations. This may result in wrongly merging spatially - temporally different activities into one activity location, or separating activities in different time periods at the same location.
2. **Sensitivity to parameter settings**: Some existing clustering methods (such as DBSCAN) require pre - set parameters (such as search radius and minimum number of points), and the selection of these parameters has a great impact on the results and is difficult to be universal in different data sets or regions.
3. **Uncertain boundaries of activity locations**: Due to the low spatial resolution of mobile phone data, the actual boundaries of activity locations may vary greatly, resulting in pre - defined spatial thresholds being difficult to apply to all cases.

#### Solutions:
To overcome these problems, the author proposes a method based on the Spatiotemporal Knowledge Graph (STKG), which can more accurately identify activity locations by considering both spatial and temporal relationships simultaneously. The specific steps are as follows:

1. **Construct STKG**: Design and construct a spatiotemporal knowledge graph that describes individual movement characteristics, infer the temporal and spatial relationships of individual stays, and convert them into a spatiotemporal graph.
2. **Apply community detection algorithm**: Use a community detection algorithm optimized for modularity (such as the Fast Unfolding algorithm) to identify stay points with dense spatiotemporal relationships, which are considered as activity locations.
3. **Verify method performance**: Conduct a case study in Shanghai to verify the effectiveness of the proposed method and compare it with two benchmark methods.

Through this method, the author hopes to improve spatial accuracy and enhance temporal consistency at the same time, so as to more accurately identify individual activity locations.