Interpretable spatio-temporal networks for modeling and forecasting societal events

Download
2019
Ertuğrul, Ali Mert
The relationships between individual activities and societal events (e.g. migrations, social movements) are complex due to the various social, temporal and spatial factors. Understanding such relationships in the context of various societal events such as street protests and opioid crisis, and forecasting these events are important as they have great impacts on public policies and supporting decision making of authorities. In this thesis, novel, spatio-temporal, deep neural networks are proposed (i) to forecast societal events and (ii) to help examine the relationships between societal events and their social and geographical contexts. The proposed models are designed to model the complex interactions between local (observed from within a location) and global (observed from all locations) activities by incorporating a new design of attentional networks. They are capable of forecasting the occurrence of future societal events and allow for interpreting what features, from which places, and how they contribute to event forecasting. Within the scope of this thesis, extensive experiments are conducted to evaluate the proposed networks on two different types of population-level societal events, namely social movements and opioid overdoses, with multiple datasets. The results indicate that the proposed models achieve superior forecasting performance than the compared methods. Also, they provide meaningful interpretations in terms of (i) what local and global activity features are more predictive, (ii) what locations have more salient contributions, and (iii) how these locations contribute to forecasting the subsequent events.

Suggestions

Social distinction and symbolic boundaries in a globalized context : leisure spaces in İstanbul
Lortoğlu, Dilber Ceren; Mutman (Yeğenoğlu), Meyda; Department of Sociology (2003)
This study focuses on the conditions and processes that strengthen social distinctions and symbolic boundaries in society. In order to fully grasp the conditions of these processes, it is not sufficient to simply study them as they are carried out on a daily basis. Therefore in this study firstly a general overview of the matter evaluated in the context of globalization. Although a variety of means are at work in constructing social distinctions and symbolic boundaries, in this study three of them have been...
Pattern recognition in bistable networks
VLADIMIR, CHINAROV; Halıcı, Uğur; Leblebicioğlu, Mehmet Kemal (1999-04-08)
Present study concerns the problem of learning, pattern recognition and computational abilities of a homogeneous network composed from coupled bistable units. An efficient learning algorithm is developed. New possibilities for pattern recognition may be realized due to the developed technique that permits a reconstruction of a dynamical system using the distributions of its attractors. In both cases the updating procedure for the coupling matrix uses the minimization of least-mean-square errors between the ...
Spatial entrapment, social mobility and education : the case of Ankara-Demetevler
Hatipoğlu, Hasan Belya; Şengül, Hüseyin Tarık; Department of Urban Policy Planning and Local Governments (2013)
Today, increasing social and economic fragmentation of the society together with important transformations in spatial structures redefine the dynamics of social mobility and education as one of the key means of defining the position of people in social hierarchy. The main objective of this thesis is to evaluate the relationship between sociospatial mobility and education by concentrating on the case study conducted in Demetevler which is one of the low-income neigbourhoods of Ankara having various dynamics ...
Personality Analysis Using Classification on Turkish Tweets
Karagöz, Pınar; Toroslu, İsmail Hakkı (2021-10-01)
According to the psychology literature, there is a strong correlation between personality traits and the linguistic behavior of people. Due to the increase in computer based communication, individuals express their personalities in written forms on social media. Hence, social media has become a convenient resource to analyze the relationship between personality traits and lingusitic behaviour. Although there is a vast amount of studies on social media, only a small number of them focus on personality predic...
Event Detection on Communities: Tracking the Change in Community Structure within Temporal Communication Networks
Aktunç, Rıza; Toroslu, İsmail Hakkı; Karagöz, Pınar (Springer, Cham, 2020-01-01)
In this work, we focus on social interactions in communities in order to detect events. There are several previous efforts for the event detection problem based on analyzing the change in the network structure in terms of the overall network features. However, in this work, event detection is considered as a problem of change detection in community structures. Particularly, communities extracted from communication network are focused on, and various versions of the community change detection methods are dev...
Citation Formats
A. M. Ertuğrul, “Interpretable spatio-temporal networks for modeling and forecasting societal events,” Thesis (Ph.D.) -- Graduate School of Natural and Applied Sciences. Information Systems., Middle East Technical University, 2019.