A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Extracting Domain Folksonomy for the Built Environment—An Automated Approach
Applicability of data models in providing semantic and harmonized representation formats of data for evaluating public opinion on the urban services depends on identifying proper attributes which is a challenging task and needs the modeling of domain knowledge. This paper presents a “bottom-up” approach for detecting key concepts to be used as data model entities while analyzing citizens’ comfort levels. In the proposed approach, comments from the end-users of urban built environment services are collected and groups of correlated keywords are automatically detected by a new fusion algorithm, which uses both co-occurrence matrix and Stanford’s dependency parsers. As a result, group representors are considered as data model attributes. The algorithm is applied to the case of pedestrian’s comfort in Canada. To evaluate the proposed approach, a total of 3130 unique, relevant, and meaningful keywords are extracted from 2000 comments collected from an urban public engagement website, called SeeClickFix. Afterward, the proposed approach is applied to form 800 groups of interrelated keywords and suggest 800 data model attributes accordingly. Reviewing the keywords suggested by the algorithm shows that most of them are meaningful and can be verified by expert’s knowledge.
Extracting Domain Folksonomy for the Built Environment—An Automated Approach
Applicability of data models in providing semantic and harmonized representation formats of data for evaluating public opinion on the urban services depends on identifying proper attributes which is a challenging task and needs the modeling of domain knowledge. This paper presents a “bottom-up” approach for detecting key concepts to be used as data model entities while analyzing citizens’ comfort levels. In the proposed approach, comments from the end-users of urban built environment services are collected and groups of correlated keywords are automatically detected by a new fusion algorithm, which uses both co-occurrence matrix and Stanford’s dependency parsers. As a result, group representors are considered as data model attributes. The algorithm is applied to the case of pedestrian’s comfort in Canada. To evaluate the proposed approach, a total of 3130 unique, relevant, and meaningful keywords are extracted from 2000 comments collected from an urban public engagement website, called SeeClickFix. Afterward, the proposed approach is applied to form 800 groups of interrelated keywords and suggest 800 data model attributes accordingly. Reviewing the keywords suggested by the algorithm shows that most of them are meaningful and can be verified by expert’s knowledge.
Extracting Domain Folksonomy for the Built Environment—An Automated Approach
Lecture Notes in Civil Engineering
Desjardins, Serge (editor) / Poitras, Gérard J. (editor) / Nik-Bakht, Mazdak (editor) / Zarei, Farzaneh (author) / Mahajan, Abhay (author) / Mock, Andrea (author) / Nik-Bakht, Mazdak (author)
Canadian Society of Civil Engineering Annual Conference ; 2023 ; Moncton, NB, Canada
2024-10-16
12 pages
Article/Chapter (Book)
Electronic Resource
English
Organizing Information in Medical Blogs Using a Hybrid Taxonomy-Folksonomy Approach
| British Library Online Contents | 2015
Crowdsourcing and the folksonomy of emergency response:The construction of a mediated subject
| BASE | 2015
TIBKAT | 4.1978,2 -