Date of Award
9-18-2018
Publication Type
Master Thesis
Degree Name
M.Sc.
Department
Computer Science
Keywords
Accessibility, Binary Rating Aggregation, binary time series, modeling, Point of Interest, Recommendation System
Supervisor
Mavromoustakos, Stephanos
Supervisor
Yuan, Xiaobu
Rights
info:eu-repo/semantics/openAccess
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Abstract
Everyone needs one or more forms of accessibility at some point in life due to age, medical conditions, accidents, etc. People with accessibility needs have the right to accessible services, as well as the right to information about accessibility at various places or Points of Interest (POI). While most popular POI recommendation services do not take accessibility into account, some of them only consider a few specific needs, such as ramp for wheelchair users. However, different users have different accessibility needs regarding the structure of the building, special aid devices, and facilities to be able to independently visit a place. The proposed system focuses on finding the personalized accessibility score for a (user, POI) pair. It can be used with other factors such as historical behavior, social influence, geographical conditions, etc. to recommend accessible places. It uses time decaying aggregate on the crowd-sourced binary rating data to find accurate approximation of current accessibility status for each accessibility criteria. Also, we propose a tunnel-based algorithm to detect the trend of binary stream data to update the rate of decay. This ensures that the calculated aggregate adapts to change in the accessibility status of the place.
Recommended Citation
Subedi, Bidur, "Accessible POI Recommendation Using Adaptive Aggregation of Binary Ratings" (2018). Electronic Theses and Dissertations. 7575.
https://scholar.uwindsor.ca/etd/7575