A THREE-STEP SPATIAL-TEMPORAL-SEMANTIC CLUSTERING METHOD FOR HUMAN ACTIVITY PATTERN ANALYSIS

How people move in cities and what they do in various locations at different times form human activity patterns. Human activity pattern plays a key role in in urban planning, traffic forecasting, public health and safety, emergency response, friend recommendation, and so on. Therefore, scholars from different fields, such as social science, geography, transportation, physics and computer science, have made great efforts in modelling and analysing human activity patterns or human mobility patterns. One of the essential tasks in such studies is to find the locations or places where individuals stay to perform some kind of activities before further activity pattern analysis. In the era of Big Data, the emerging of social media along with wearable devices enables human activity data to be collected more easily and efficiently. Furthermore, the dimension of the accessible human activity data has been extended from two to three (space or space-time) to four dimensions (space, time and semantics). More specifically, not only a location and time that people stay and spend are collected, but also what people “say” for in a location at a time can be obtained. The characteristics of these datasets shed new light on the analysis of human mobility, where some of new methodologies should be accordingly developed to handle them. Traditional methods such as neural networks, statistics and clustering have been applied to study human activity patterns using geosocial media data. Among them, clustering methods have been widely used to analyse spatiotemporal patterns. However, to our best knowledge, few of clustering algorithms are specifically developed for handling the datasets that contain spatial, temporal and semantic aspects all together. In this work, we propose a three-step human activity clustering method based on space, time and semantics to fill this gap. One-year Twitter data, posted in Toronto, Canada, is used to test the clustering-based method. The results show that the approximate 55% spatiotemporal clusters distributed in different locations can be eventually grouped as the same type of clusters with consideration of semantic aspect.

Download Full-text

VISUALISATION OF SOCIO-SPATIAL ISOLATION BASED ON HUMAN ACTIVITY PATTERNS AND SOCIAL NETWORKS IN SPACE-TIME

Tijdschrift voor Economische en Sociale Geografie ◽

10.1111/j.1467-9663.2010.00649.x ◽

2011 ◽

Vol 102 (4) ◽

pp. 468-485 ◽

Cited By ~ 56

Author(s):

JAE YONG LEE ◽

MEI-PO KWAN

Keyword(s):

Social Networks ◽

Human Activity ◽

Activity Patterns ◽

Space Time ◽

Spatial Isolation ◽

Human Activity Patterns

Download Full-text

Understanding human activity patterns based on space-time-semantics

ISPRS Journal of Photogrammetry and Remote Sensing ◽

10.1016/j.isprsjprs.2016.08.008 ◽

2016 ◽

Vol 121 ◽

pp. 1-10 ◽

Cited By ~ 19

Author(s):

Wei Huang ◽

Songnian Li

Keyword(s):

Human Activity ◽

Activity Patterns ◽

Space Time ◽

Time Semantics ◽

Human Activity Patterns

Download Full-text

A NEW HIERARCHICAL CLUSTERING APPROACH FOR SPARSE MOBILE PHONE TRAJECTORIES

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprs-archives-xlii-4-697-2018 ◽

2018 ◽

Vol XLII-4 ◽

pp. 697-701

Author(s):

W. Wang ◽

Z. Luan ◽

B. He ◽

X. Li ◽

D. Zhang ◽

...

Keyword(s):

Mobile Phone ◽

Hierarchical Clustering ◽

Human Activity ◽

Urban Space ◽

Activity Pattern ◽

Activity Patterns ◽

Social Activity ◽

Spatial Uncertainty ◽

Clustering Approach ◽

Human Activity Patterns

Abstract. Understanding the pattern of human activities benefits both the living service providing for the public and the policy-making for urban planners. The development of location-aware technology enables us to acquire large volume individual trajectories with different spatial and temporal resolution, such as GPS trajectories, mobile phone positioning data, social media check-in data, Wifi, and Bluetooth. However, the highest population penetrated mobile phone positioning trajectories are hard to infer human activity pattern directly, because of the sparsity in both space and time. This article presents a hierarchical clustering approach by using the move and stay sequences inferred from spare mobile phone trajectories to uncover the hidden human activity pattern. Personal stays at some places and following moves are first extracted from mobile phone trajectories, considering the spatial uncertainty of position. The similarity of trajectories is measured with a new indicator defined by the area of a spatial-temporal polygon bound with normalized trajectories. Finally, a hierarchical clustering method is developed to group trajectories with similar stay-move chains from the bottom to the top. The obtained clusters are analyzed to identify human activity patterns. An experiment with mobile phone users’ one-day trajectories in Shenzhen, China was conducted to test the performance of the proposed clustering approach. The results indicate all used trajectories are classified into 10 clusters representing typical daily activity patterns from the simple home-staying mode to complex home-working-social activity daily cycles. This study not only unravels the hidden activity patterns behind massive sparse trajectories but also deepens our understanding of the interaction of human activity and urban space.

Download Full-text

Modeling intra-urban human activity patterns using crowdsourcing GPS and Geosocial Media Data

10.32920/ryerson.14647764.v1 ◽

2021 ◽

Author(s):

Wei Huang

Keyword(s):

Predictive Model ◽

Human Activity ◽

Activity Pattern ◽

Activity Patterns ◽

Urban Systems ◽

Significant Activity ◽

Human Movements ◽

Human Activity Patterns ◽

Media Data ◽

The Way

The way people live in cities forms human activity patterns, which affects how urban systems work. Therefore, it is essential to understand human activity patterns, where precise prediction of human movements and mechanistic modelling of human activity patterns are the two keys. Most of existing work on prediction of human movements cannot deal with activity changes, leading to a negative impact on the predictive accuracy. Furthermore, the majority of current work on modelling human activity patterns are mainly researched from spatiotemporal perspectives, but the motivation behind is usually being neglected, which is crucial to understanding activity changes. The objective of this study is to develop models and methods to better understand human activity patterns using crowdsourcing and geosocial media data. Thus, in this thesis, a method is first developed to detect activity changes, based on which a Markov chain-based model is developed to predict human movements. Then, semantics is introduced to uncover the motivation associated with the corresponding spatiotemporal patterns, which can infer what people do and discuss in a location at a specific time. Finally, human activity patterns are modelled from both spatiotemporal and semantic perspectives. A 6-year GPS dataset of human movement in Beijing, China was used to evaluate the proposed predictive model. The results show that the predictive model can yield accurate prediction of the movement for those users who have significant activity changes (with R2 improved from 0.295 to 0.762). A whole-year geo-tagged tweets posted within Toronto, Canada was acquired to analyze human activity patterns. A network model was finally created by the proposed approach to represent human activity patterns. The experimental findings demonstrate that most of the individuals (61%) have a regular activity pattern, while only a small number of people (10%) have a different activity pattern from the mass. With the inclusion of semantic information together with the spatiotemporal data as well as detecting the activity changes, such an approach can enhance the capability of human mobility and activity modelling, and thus pave the way for a more mechanistic understanding of how urban systems are being shaped, as well as how their sub-systems/components interact.

Download Full-text

Modeling intra-urban human activity patterns using crowdsourcing GPS and Geosocial Media Data

10.32920/ryerson.14647764 ◽

2021 ◽

Author(s):

Wei Huang

Keyword(s):

Predictive Model ◽

Human Activity ◽

Activity Pattern ◽

Activity Patterns ◽

Urban Systems ◽

Significant Activity ◽

Human Movements ◽

Human Activity Patterns ◽

Media Data ◽

The Way

The way people live in cities forms human activity patterns, which affects how urban systems work. Therefore, it is essential to understand human activity patterns, where precise prediction of human movements and mechanistic modelling of human activity patterns are the two keys. Most of existing work on prediction of human movements cannot deal with activity changes, leading to a negative impact on the predictive accuracy. Furthermore, the majority of current work on modelling human activity patterns are mainly researched from spatiotemporal perspectives, but the motivation behind is usually being neglected, which is crucial to understanding activity changes. The objective of this study is to develop models and methods to better understand human activity patterns using crowdsourcing and geosocial media data. Thus, in this thesis, a method is first developed to detect activity changes, based on which a Markov chain-based model is developed to predict human movements. Then, semantics is introduced to uncover the motivation associated with the corresponding spatiotemporal patterns, which can infer what people do and discuss in a location at a specific time. Finally, human activity patterns are modelled from both spatiotemporal and semantic perspectives. A 6-year GPS dataset of human movement in Beijing, China was used to evaluate the proposed predictive model. The results show that the predictive model can yield accurate prediction of the movement for those users who have significant activity changes (with R2 improved from 0.295 to 0.762). A whole-year geo-tagged tweets posted within Toronto, Canada was acquired to analyze human activity patterns. A network model was finally created by the proposed approach to represent human activity patterns. The experimental findings demonstrate that most of the individuals (61%) have a regular activity pattern, while only a small number of people (10%) have a different activity pattern from the mass. With the inclusion of semantic information together with the spatiotemporal data as well as detecting the activity changes, such an approach can enhance the capability of human mobility and activity modelling, and thus pave the way for a more mechanistic understanding of how urban systems are being shaped, as well as how their sub-systems/components interact.

Download Full-text