|Ph.D Student||Galon Binyamin|
|Subject||Bridging Social Networks and Spatial Data|
|Department||Department of Civil and Environmental Engineering||Supervisors||Professor Emeritus Yerach Doytsher|
|Dr. Yaron Kanza|
|Full Thesis text|
Cellular phones and GPS-based navigation systems allow recording the location history of users, to find the places that users visit and the routes along which they travel.
In addition, the continually increasing usage of the Internet and especially online social network (e.g. Facebook, Instagram, Tweeter, etc.), facilitates collection of data about people and their relationships.
This provides associations between users, social relationships and geographic entities. Data that include both social relationships and location history of people are referred to as socio-spatial data.
Queries over socio-spatial data retrieve information on users, corresponding to their location history, and retrieve information on geographical entities (locations) corresponding to the users who visit these places.
In this research, we present a novel graph model for socio-spatial networks that store information on: (1) users and their relationships, (2) geospatial entities (e.g. houses, roads, neighborhoods, etc.) and their topological relations, (3) user visits to locations and the traveled routes to reach that location.
We present a query language that consists of graph traversal operations, aiming at facilitating the formulation of queries, and we show how queries over the network can be evaluated efficiently.
We also show how social-based route recommendation can be implemented using our query language. Social-based route recommendation systems recommend routes between two locations based on user similarity, and not based on the shortest path. For example, say a user tends to avoid left turns, then a social route recommendation system could recommend routes based on the travel history of other users also avoiding left turns.
Another aspect of socio-spatial data is the emotions users feel when visiting a location. There is often a strong correlation between the environment and the way people feel, e.g., the emotions associated with a hospital are typically very different from those associated with an amusement park or a promenade.
In the last part of the research we introduce the concept of emotion maps, which aim to represent and depict interrelationships between emotions and geographic locations. Such maps can provide answers to various questions about how people feel at various places or at different times of the day. They can facilitate a search for places where people express a certain emotion.
We introduce a new approach for creating emotion maps from a large collection of geotagged social-media posts. We discuss potential usages of such maps. We present a model to query and utilize emotion maps and we demonstrate the creation of emotion maps by applying emotion analysis to millions of geotagged tweets.