{"id":{"repo_id":"buffalo","oai_identifier":"oai:ubir.buffalo.edu:10477/79360"},"canonical_url":"https://search.dev.ndltd.org/etd/buffalo/oai:ubir.buffalo.edu:10477/79360","repository":{"repo_id":"buffalo","name":"Buffalo","base_url":"https://ubir.buffalo.edu/oai/request"},"display":{"title":"Augmenting Household Travel Survey and Travel Behavior Analysis Using Large-Scale Social Media Data and Smartphone GPS Data","abstract":"Ph.D.","abstract_html":"Ph.D.","abstract_has_math":false,"creators":["Cui, Yu; 0000-0001-7916-1605"],"institution":"State University of New York at Buffalo","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["He, Qing","Civil, Structural and Environmental Engineering"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2019,"date_issued":"2019-04-04T20:30:50Z","date_published":"2019-04-04T20:30:50Z","updated_at":"2026-07-27T19:05:16Z","subjects":["transportation"],"languages":["eng"],"rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/10477/79360","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["He, Qing","Civil, Structural and Environmental Engineering"]},{"key":"dc:creator","label":"Author","values":["Cui, Yu; 0000-0001-7916-1605"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2019-04-04T20:30:50Z","2019","2018-12-24 00:50:20"]},{"key":"dc:publisher","label":"Institution","values":["State University of New York at Buffalo"]},{"key":"dc:type","label":"Dc Type","values":["Text","Dissertation"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["transportation"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/10477/79360"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Ph.D.","Household travel survey data is a critical input to travel behavior modeling, and it also can be used to generate trip chains for activity-based traffic simulation. However, traditional household travel survey usually collects a very short period of travel (e.g. 1 day). As a result, the individual’s longitudinal travel behavior can be barely captured with this method. A growing body of literature suggests that longer data collection periods are warranted to provide improved data for modeling purposes and understanding travel variations. Therefore, there is a pressing need to conduct longer periods of data collection and devise a new approach which avoids the possibly imposed respondent burden and survey costs. With emerging information and communication technology (ICT) tools, the collection of passive datasets for travelers’ real-time information becomes available. High penetration of smartphones guarantees that collecting GPS trajectory becomes trivial. Smartphone GPS survey apps have emerged to be a popular tool for conducting household travel surveys. However, existing smartphone surveys are still relatively limited in terms of study periods due to aforementioned reasons. Recently, social media has proved to be a valuable data source for conducting research in the transportation domain. It can provide an individual’s years of posts and check-in geo-locations for prediction of trip purpose, which is crucial to travel behavior modeling and travel demand estimation for transportation planning and investment decisions. However, many challenges still exist when one adopts social media data. A well-known challenge is that social media users’ population cannot represent the real population. To fully utilize social media data, it is essential to correct sampling biases of social media data beforehand. This dissertation paves a new way to travel survey data collection and travel behavior analysis. It consists of four major components. The first study is about uncovering different kinds of travel behavior patterns based on traditional travel survey data. Jaccard similarity coefficient is employed to help to construct community social network for community detection from activity matrix. A deep learning approach with convolution neural network is employed to classify travelers into corresponding groups according to their activity maps. The accuracy of classification reaches up to 97%. The second study aims to predict the current and next trip purpose, which is an essential attribute in travel behavior research. Social media and Google Places data are employed in this study. A Bayesian neural network (BNN) is employed to implemented to model the trip dependence within each individual’s daily trip chain and infer the trip purpose. Further, to tackle the computational challenge in BNN, Elastic Net is used for feature selection before the classification task. Comparing with traditional models, it is found that Google Places and Twitter information can greatly improve the overall accuracy of prediction. Moreover, this information also helps to improve the accuracy of each trip purpose, especially eating out, personal business, recreation, and shopping activities. The objective of the third component is to develop an approach to resample social media data in order to reduce biases and errors through estimation of socio-demographics. Several machine learning models are proposed for predicting socio-demographics, including gender, age, ethnicity and education levels. Afterward, this study resamples social media data and compares the results with the 2009 California Household Travel Survey data. The resampled data shows comparable characteristics to the survey data. Moreover, since social media is a kind of long-term data, it shows several advantages in research over survey data. This research sheds light on tackling sampling bias issues when social media data is used for travel behavior analysis.The fourth study shapes a sustainable and long-term travel survey with 7-month low-frequency smartphone GPS data with imperfect activity information. The essential goal is to develop a daily synthetic trip chain simulator. This research develops a new probabilistic method to handle imperfect activity data, and three different levels of trip chain generation models are proposed. The first model handles only known activities, the second model treats all unknown activities as a single category, and the third one models each unknown location separately. These models are able to generate trip chains in different levels of details for activity-based traffic simulator."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Augmenting Household Travel Survey and Travel Behavior Analysis Using Large-Scale Social Media Data and Smartphone GPS Data"]}]}],"canonical_facts":{"dc:contributor":["He, Qing","Civil, Structural and Environmental Engineering"],"dc:creator":["Cui, Yu; 0000-0001-7916-1605"],"dc:date":["2019-04-04T20:30:50Z","2019","2018-12-24 00:50:20"],"dc:description":["Ph.D.","Household travel survey data is a critical input to travel behavior modeling, and it also can be used to generate trip chains for activity-based traffic simulation. However, traditional household travel survey usually collects a very short period of travel (e.g. 1 day). As a result, the individual’s longitudinal travel behavior can be barely captured with this method. A growing body of literature suggests that longer data collection periods are warranted to provide improved data for modeling purposes and understanding travel variations. Therefore, there is a pressing need to conduct longer periods of data collection and devise a new approach which avoids the possibly imposed respondent burden and survey costs. With emerging information and communication technology (ICT) tools, the collection of passive datasets for travelers’ real-time information becomes available. High penetration of smartphones guarantees that collecting GPS trajectory becomes trivial. Smartphone GPS survey apps have emerged to be a popular tool for conducting household travel surveys. However, existing smartphone surveys are still relatively limited in terms of study periods due to aforementioned reasons. Recently, social media has proved to be a valuable data source for conducting research in the transportation domain. It can provide an individual’s years of posts and check-in geo-locations for prediction of trip purpose, which is crucial to travel behavior modeling and travel demand estimation for transportation planning and investment decisions. However, many challenges still exist when one adopts social media data. A well-known challenge is that social media users’ population cannot represent the real population. To fully utilize social media data, it is essential to correct sampling biases of social media data beforehand. This dissertation paves a new way to travel survey data collection and travel behavior analysis. It consists of four major components. The first study is about uncovering different kinds of travel behavior patterns based on traditional travel survey data. Jaccard similarity coefficient is employed to help to construct community social network for community detection from activity matrix. A deep learning approach with convolution neural network is employed to classify travelers into corresponding groups according to their activity maps. The accuracy of classification reaches up to 97%. The second study aims to predict the current and next trip purpose, which is an essential attribute in travel behavior research. Social media and Google Places data are employed in this study. A Bayesian neural network (BNN) is employed to implemented to model the trip dependence within each individual’s daily trip chain and infer the trip purpose. Further, to tackle the computational challenge in BNN, Elastic Net is used for feature selection before the classification task. Comparing with traditional models, it is found that Google Places and Twitter information can greatly improve the overall accuracy of prediction. Moreover, this information also helps to improve the accuracy of each trip purpose, especially eating out, personal business, recreation, and shopping activities. The objective of the third component is to develop an approach to resample social media data in order to reduce biases and errors through estimation of socio-demographics. Several machine learning models are proposed for predicting socio-demographics, including gender, age, ethnicity and education levels. Afterward, this study resamples social media data and compares the results with the 2009 California Household Travel Survey data. The resampled data shows comparable characteristics to the survey data. Moreover, since social media is a kind of long-term data, it shows several advantages in research over survey data. This research sheds light on tackling sampling bias issues when social media data is used for travel behavior analysis.The fourth study shapes a sustainable and long-term travel survey with 7-month low-frequency smartphone GPS data with imperfect activity information. The essential goal is to develop a daily synthetic trip chain simulator. This research develops a new probabilistic method to handle imperfect activity data, and three different levels of trip chain generation models are proposed. The first model handles only known activities, the second model treats all unknown activities as a single category, and the third one models each unknown location separately. These models are able to generate trip chains in different levels of details for activity-based traffic simulator."],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/10477/79360"],"dc:language":["eng"],"dc:publisher":["State University of New York at Buffalo"],"dc:rights":["Users of works found in University at Buffalo Institutional Repository (UBIR) are responsible for identifying and contacting the copyright owner for permission to reuse. University at Buffalo Libraries do not manage rights for copyright-protected works and cannot assist with permissions.","Copyright retained by author."],"dc:subject":["transportation"],"dc:title":["Augmenting Household Travel Survey and Travel Behavior Analysis Using Large-Scale Social Media Data and Smartphone GPS Data"],"dc:type":["Text","Dissertation"]},"updated_at":"2026-07-27T19:05:16Z"}