Heri-Graphs: A Dataset Creation Framework for Multi-Modal Machine Learning on Graphs of Heritage Values and Attributes with Social Media
Abstract
:1. Introduction
- Domain-specific multi-modal attributed graphsocial network datasets about heritage values and attributes (or more precisely, the values and attributes conveyed by the public to urban cultural heritage) are collected and structured with user-generated content from the social media platform Flickr in three cities (Amsterdam, Suzhou, and Venice) containing UNESCO World Heritage properties, which could benefit the knowledge documentation and mapping for heritage and urban studies, aiming at a more inclusive heritage management process;
- Several pre-trained machine learning and deep learning models have been extensively applied and tested for generating multi-modal features and [pseudo-]labels with full mathematical formulations as its problem definition, providing a reproducible methodological framework that could also be tested in other cases worldwide;
- Multi-graphs have been constructed to reflect the temporal, spatial, and social relationships among the data samples of collected user-generated content, ready to be further tested on several provisional tasks with both scientific relevance for Graph-based Multi-modal Machine Learning and Social Network research, and societal interests for Urban Studies, Urban Data Science, and Heritage Studies.
2. Materials and Methods
2.1. General Framework
2.2. Selection of Case Studies
2.3. Data Collection and Pre-Processing
2.4. Multi-Modal Feature Generation
2.4.1. Visual Features
2.4.2. Textual Features
2.4.3. Contextual Features
2.5. Pseudo-Label Generation
2.5.1. Heritage Values as OUV Selection Criteria
2.5.2. Heritage Attributes as Depicted Scenery
2.6. Multi-Graph Construction
2.6.1. Temporal Links
2.6.2. Social Links
2.6.3. Spatial Links
3. Analyses as Qualitative Inspection
3.1. Sample-Level Analyses of Datasets
3.1.1. Generated Visual and Textual Features
3.1.2. Pseudo-Labels for Heritage Values and Attributes
3.2. Graph-Level Analyses of Datasets
3.2.1. Back-End Geographical Network
3.2.2. Multi-Graphs and Sub-Graphs of Contextual Information
4. Discussion
4.1. Provisional Tasks for Urban Data Science
ID | Problem Definition | Type of Task | As a Machine Learning/Social Network Analysis Problem | As an Urban/Heritage Study Question |
---|---|---|---|---|
0 | Image Classification (semi-supervised) | Using visual features to infer categories originally induced from (possibly missing) texts with co-training [112] in few-shot learning settings [113]. | As the latest advances in heritage value assessment have been discovering the added value of inspecting texts [4], can values also be seen and retrieved from scenes of images? | |
1 | Text Classification (semi-supervised) | Using textual features to infer categories originally induced from images possibly with attention mechanisms [75]. | How to relate the textual descriptions to certain heritage attributes [28]? Are there crucial hints other than appeared nouns? | |
2 | Multi-modal Classification (semi-supervised) | Using multi-modal (multi-view) features to make inference, either with training joint representations or by making early and/or late fusions [13,112]. | How can heritage values and attributes be jointly inferred from the combined information of both visual scenes and textual expressions [26]? How can they complement each other? | |
3 | Node Classification (semi-supervised) | Test-beds for different graph filters such as Graph Convolution Networks [50] and Graph Attention Networks [114]. | How can the contextual information of a post contribute to the inference of its heritage values and attributes? What is the contribution of time, space, and social relations [115]? | |
4 | Link Prediction and Recommendation System (semi-supervised) | Test-beds for link prediction algorithms [116] considering current graph structure and node features. What is the probability that other links also should exist? | Considering the similarity of posts, would there be heritage values and attributes that also suit the interest of another user, fit another location, and/or reflect another period of time [117]? | |
5 | Graph Coarsening (unsupervised) | Test-beds for graph pooling [44] and graph partitioning [106] algorithms to generate coarsened graphs [118] in different resolutions. | How can we summarize, aggregate, and eventually visualize the large-scale information from the social media platforms based on their contents and contextual similarities [30]? | |
6 | Graph Classification (supervised) | Test-beds for graph classification algorithms [119] when more similar datasets have been collected and constructed in more case study cities. | Can we summarize the social media information of any city with World Heritage property so that the critical heritage values and attributes could be directly inferred [25]? | |
7 | Image/Text Generation (supervised) | Using multi-modal features to generate the missing and/or unfit images and/or textual descriptions, probably with Generative Adversarial Network [120]. | How can a typical image and/or textual description of certain heritage values and attributes at a certain location in a certain time by a certain type of user in a specific case study city be queried or even generated [28]? | |
8 | Attributed Multi-Graph Embedding (self-supervised) | Respectively generating a universal embedding and a context-specific embedding for each type of links in the multi-dimensional network [121], probably with random walks on graphs. | How are heritage values and attributes distributed and diffused in different contexts? Is the First Law of Geography [32] still valid in the specific social, temporal and spatial graphs? | |
9 | Dynamic Prediction (self-supervised) | Given the current graph structure and its features stamped with time steps, how shall it further evolve in the next time steps [36,122]? | How are the current expressions of heritage values and attributes in a city influencing the emerging post contents, the tourist behaviours, and the planning decision making [9,37]? |
4.2. Limitations and Future Steps
5. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Abbreviations
Acc | Accuracy |
AMS | Data of Amsterdam, The Netherlands |
API | Application Programming Interface |
BC-SVM | Bagging Classifier with SVMs as the internal base estimators |
BERT | Bidirectional Encoder Representations from Transformers |
CNN | Convolutional Neural Network |
CC | Connected Components |
CV | Cross-Validation |
GNB | Gaussian Naive Bayes |
HA | Heritage Attributes |
HUL | Historic Urban Landscape |
HV | Heritage Values |
KNN | K-Nearest Neighbours |
ML | Machine Learning |
MLP | Multi-layer Perceptron |
MML | Multi-modal Machine Learning |
NLP | Natural Language Processing |
OUV | Outstanding Universal Value |
RF | Random Forest |
SDG | Sustainable Development Goal |
SNA | Social Network Analysis |
SUZ | Data of Suzhou, China |
SVM | Support Vector Machine |
UGC | User-Generated Content |
ULMFiT | Universal Language Model Fine-tuning |
UNESCO | The United Nations Educational, Scientific and Cultural Organization |
VEN | Data of Venice, Italy |
VEN-XL | The extra-large version of Venice data |
w. | with |
WH | World Heritage |
WHL | World Heritage List |
Appendix A. Details of Collecting the Raw Dataset
Appendix B. Details for Machine Learning Models
Appendix B.1. MLP
Appendix B.2. KNN
Appendix B.3. GNB
Appendix B.4. SVM
Appendix B.5. RF
Appendix B.6. Bagging
Appendix B.7. Voting
Appendix B.8. Stacking
Appendix C. Nomenclature
Symbol | Data Type/Shape | Description |
---|---|---|
Matrix of Boolean | The adjacency matrix of all post nodes in the set that have at least one link connecting them as a composed simple graph. | |
Matrix of Float , | The weighted adjacency matrix of each of the three sub-graphs of the multi-graph , “(*)” represents one of the link types in {TEM, SOC, SPA}. | |
Matrix of Boolean | The adjacency matrix of all unique users marking their direct friendship which also included the relationship among themselves. | |
Matrix of Float | The weighted adjacency matrix of all unique users marking their mutual interest in terms of the Jaccard Index of the public groups that they follow. | |
Float scalars | Parameters adjusting the weights of linear combination in relationship matrices and . | |
Float scalar | The threshold to define mutual interest of two users as the Jaccard Index of public groups. | |
Float Scalar | The Chi-square statistics of two distributions. | |
Object Tuples | The tuple of all raw data (image, sentences, user ID, timestamp, and geo-location) from one sample point. | |
Float Scalar | The Kullback–Leibler (KL) divergence of two distributions. | |
Float Scalar | An arbitrary small number to avoid zero-division. | |
Matrix of Integers and Floats | The face recognition result of an image sample in terms of the number of faces detected , the model confidence for the prediction , and the proportion of total area of bounding boxes of detected faces to the total area of images . | |
Undirected weighted graph | The complete spatial network in a city weighted by the travel time with all sorts of transportation between spatial nodes. | |
G | Undirected weighted graph | The spatial network in a city weighted by the travel time between spatial nodes (no more than 20 min) that have at least one sample posted near them. |
Weighted multi-graph | The graph including the temporal, social, and spatial links among the post nodes from set , weighted by the respective connection strengths . | |
Undirected weighted graph , | The sub-graph of the multi-graph , while “(*)” represents one of the link types in {TEM, SOC, SPA}. | |
Matrix of Floats | The last hidden layer for [CLS] token of BERT model pre-trained on WHOSe_Heritage. | |
Matrix of Floats | The last hidden layer of ResNet-18 model pre-trained on Places365. | |
Integer Indices | The index of samples in the dataset of one case city. | |
Tensor of Integers within of size or | The raw image data of one sample post with RGB channels. | |
Matrix of Boolean | The diagonal identity matrix marking the identity of unique users in . | |
Integer Indices | The index of users in the ordered set of all unique users from one case city. | |
k | Integer Indices | The index of timestamps in the ordered set of all unique timestamps from one case city. |
K | Integer | The sample size (number of posts) collected in one case city. |
Matrix of Floats | The confidence indicator matrix for heritage attributes labels including the top-n confidence and agreement between VOTE and STACK models. | |
Matrix of Floats | The confidence indicator matrix for heritage values labels including the top-n confidence and agreement between BERT and ULMFiT models. | |
Integer Indices | The index of nodes in the ordered set V of all spatial nodes from one case city. | |
Tuple of Floats | The geographical coordinate of latitude () and longitude () as location of one sample. | |
Matrix of logit vectors | The last softmax layer of ResNet-18 model pre-trained on SUN predicting scene attributes. | |
Matrix of logit vectors | The last softmax layer of ResNet-18 model pre-trained on Places365 predicting scene categories. | |
A set of objects | The set of machine learning models used to train classifiers on Tripoli data. | |
Matrix of Boolean | The language detection result of the original language appearance of the sentences in each sample, in terms of English , local language , and other languages . | |
Matrix of Float , | The embedding matrices of each of the samples to a N-dimensional vector based on the general structure of the multi-graph and the specific types of links. | |
Set of Strings or Empty Set | The processed textual data as a set of individual sentences that have a valid semantic meaning and have been translated into English. | |
Boolean Matrix | The one-hot embedding matrix of the samples corresponding to the geo-node set V. | |
Matrix of Float | A matrix marking the spatial closeness of all the unique spatial nodes from set V that can be reached within 20 min. | |
An ordered Set | The ordered set of all unique timestamps from one case city. | |
Timestamp | A timestamp in the ordered set of all unique timestamps. | |
Timestamp | A timestamp indexed with sample ID in the ordered set of all unique timestamps. | |
Boolean Matrix | The one-hot embedding matrix of the samples corresponding to the timestamp set . | |
Matrix of Float | A matrix marking the temporal similarity of all the unique timestamps from set . | |
An ordered Set | The ordered set of all unique users from one case study city. | |
User ID Object | An instance of user in the ordered set of all unique users. | |
User ID Object | An instance of user indexed with sample ID in the ordered set of all unique users. | |
Boolean Matrix | The one-hot embedding matrix of the samples corresponding to the user set . | |
Matrix of Float | A matrix marking the social similarity of all the unique users from set , as a linear combination of identity matrix and adjacency matrices . | |
V | A set of nodes | The set of all the spatial nodes that have at least one sample posted near them. |
Spatial node | A node in the set V of all spatial nodes that have at least one sample posted near them. | |
A set of nodes | The set of all nodes of posts in one case city. | |
Post/Sample node | A node in the set of all nodes of posts in one case city. | |
Vector of Float | The weight vector of spatial network G and post graphs , these weights are directly interchangeable with the adjacency matrices. | |
Matrix of Floats and Integers | The final visual feature concatenating the hidden layer , the face detection results , the filtered top-5 scene prediction , and the filtered top-10 attribute prediction . | |
Matrix of Floats and Integers | The final textual feature concatenating the hidden layer of BERT on [CLS] token, and the original language detection results . | |
Matrix of Floats | The final generated label of heritage attributes on 9 depicted scenes, as the average of prediction from VOTE and STACK models. | |
Matrix of Floats | The final generated label of heritage values on 10 OUV selection criteria and an additional negative class, as the average of prediction from BERT and ULMFiT models. |
Symbol | Data Type/Shape | Description |
---|---|---|
Function outputting a set of floats or objects | The set of largest n elements of any float vector . | |
Function inputting a sentence/paragraph or a batch of sentences/paragraphs, outputting a vector or a matrix of vectors | The pre-trained uncased BERT model fine-tuned on WHOSe_Heritage with the model parameters that can process some textual inputs into the 768-dimensional hidden output vector of the [CLS] token. | |
Function inputting a tensor or a batch of tensors, outputting three vectors or three matrices of vectors | The ResNet-18 model pre-trained on Places365 dataset with the model parameters that can process the image tensor into the predicted vectors of scenes , predicted vectors of attributes , and the last hidden layer . | |
Function inputting a sentence/paragraph or a batch of sentences/paragraphs, outputting a vector or a matrix of vectors | The end-to-end pre-trained uncased BERT model fine-tuned on WHOSe_Heritage with the model parameters together with the MLP classifiers that can process some textual inputs into the logit prediction vector of 11 heritage value classes concerning OUV. | |
Function inputting a sentence/paragraph or a batch of sentences/paragraphs, outputting a vector or a matrix of vectors | The end-to-end pre-trained ULMFiT model fine-tuned on WHOSe_Heritage with the model parameters together with the MLP classifiers that can process some textual inputs into the logit prediction vector of 11 heritage value classes concerning OUV. | |
Function inputting a vector or a batch of vectors, outputting a vector or a matrix of vectors | The ensemble Voting Classifier with model parameter of machine learning models from with their respective model parameters , which processes the visual feature vector into the logit prediction vector of 9 heritage attribute classes concerning depicted scenes. | |
Function inputting a vector or a batch of vectors, outputting a vector or a matrix of vectors | The ensemble Stacking Classifier with model parameter of machine learning models from with their respective model parameters , which processes the visual feature vector into the logit prediction vector of 9 heritage attribute classes concerning depicted scenes. | |
Function outputting an ordered set of objects | The set of public groups that are followed by user . | |
Function outputting a non-negative float | The Jaccard Index of any two sets as the cardinality of the intersection of the two sets over that of the union of them. | |
Function outputting a float | The largest element of any float vector . | |
Function both inputting and outputting a logit vector | The activation filter to keep the top-n entries of any logit vector and smooth all the others entries based on the total confidence (sum) of top-n entries. |
Appendix D. Definition of Categories for Heritage Values and Attributes
Criterion | Focus | Definition |
---|---|---|
(i) | Masterpiece | To represent a masterpiece of human creative genius; |
(ii) | Values/Influence | To exhibit an important interchange of human values, over a span of time or within a cultural area of the world, on developments in architecture or technology, monumental arts, town-planning or landscape design; |
(iii) | Testimony | To bear a unique or at least exceptional testimony to a cultural tradition or to a civilization which is living or which has disappeared; |
(iv) | Typology | To be an outstanding example of a type of building, architectural or technological ensemble or landscape which illustrates (a) significant stage(s) in human history; |
(v) | Land-Use | To be an outstanding example of a traditional human settlement, land-use, or sea-use which is representative of a culture (or cultures), or human interaction with the environment especially when it has become vulnerable under the impact of irreversible change; |
(vi) | Associations | To be directly or tangibly associated with events or living traditions, with ideas, or with beliefs, with artistic and literary works of outstanding universal significance; |
(vii) | Natural Beauty | To contain superlative natural phenomena or areas of exceptional natural beauty and aesthetic importance; |
(viii) | Geological Process | To be outstanding examples representing major stages of earth’s history, including the record of life, significant on-going geological processes in the development of landforms, or significant geomorphic or physiographic features; |
(ix) | Ecological Process | To be outstanding examples representing significant on-going ecological and biological processes in the evolution and development of terrestrial, fresh water, coastal and marine ecosystems and communities of plants and animals; |
(x) | Bio-diversity | To contain the most important and significant natural habitats for in situ conservation of biological diversity, including those containing threatened species of outstanding universal value from the point of view of science or conservation. |
Attribute | Type | Definition |
---|---|---|
Monuments and Buildings | Tangible | The exterior of a whole building, structure, construction, edifice, or remains that host(ed) human activities, storage, shelter or other purpose; |
Building Elements | Tangible | Specific elements, details, or parts of a building, which can be constructive, constitutive, or decorative; |
Urban Form Elements | Tangible | Elements, parts, components, or aspects of/in the urban landscape, which can be a construction, structure, or space, being constructive, constitutive, or decorative; |
Urban Scenery | Tangible | A district, a group of buildings, or specific urban ensemble or configuration in a wider (urban) landscape or a specific combination of cultural and/or natural elements; |
Natural Features and Landscape Scenery | Tangible | Specific flora and/or fauna, such as water elements of/in the historic urban landscape produced by nature, which can be natural and/or designed; |
Interior Scenery | Tangible/Intangible | The interior space, structure, construction, or decoration that host(ed) human activity, showing a specific (typical, common, special) use or function of an interior place or environment; |
People’s Activity and Association | Intangible | Human associations with a place, element, location, or environment, which can be shown with the activities therein; |
Gastronomy | Intangible | The (local) food-related practices, traditions, knowledge, or customs of a community or group, which may be associated with a community or society and/or their cultural identity or diversity; |
Artifact Products | Intangible | The (local) artifact-related practices, traditions, knowledge, or customs of a community or group, which may be associated with a community or society and/or their cultural identity or diversity. |
Appendix E. Multi-Graph Visualization
References
- UNESCO. Convention Concerning the Protection of the World Cultural and Natural Heritage; Technical Report; UNESCO: Paris, France, 1972. [Google Scholar]
- UNESCO. Operational Guidelines for the Implementation of the World Heritage Convention; Technical Report; UNESCO World Heritage Centre: Paris, France, 2008. [Google Scholar]
- Pereira Roders, A. Re-Architecture: Lifespan Rehabilitation of Built Heritage. Ph.D. Thesis, Technische Universiteit Eindhoven, Eindhoven, The Netherlands, 2007. [Google Scholar] [CrossRef]
- Tarrafa Silva, A.; Pereira Roders, A. The cultural significance of World Heritage cities: Portugal as case study. In Proceedings of the 2nd International Conference on Heritage and Sustainable Development (HERITAGE 2010), Evora, Portugal, 22–26 June 2010; pp. 255–263. [Google Scholar] [CrossRef]
- Veldpaus, L. Historic Urban Landscapes: Framing the Integration of Urban and Heritage Planning in Multilevel Governance. Ph.D. Thesis, Technische Universiteit Eindhoven, Eindhoven, The Netherlands, 2015. [Google Scholar]
- Rakic, T.; Chambers, D. World Heritage: Exploring the Tension Between the National and the ‘Universal’. J. Herit. Tour. 2008, 2, 145–155. [Google Scholar] [CrossRef]
- Bonci, A.; Clini, P.; Martin, R.; Pirani, M.; Quattrini, R.; Raikov, A. Collaborative Intelligence Cyber-physical System for the Valorization and Re-use of Cultural Heritage. J. Inf. Technol. Constr. 2018, 23, 305–323. [Google Scholar]
- Pereira Roders, A. The Historic Urban Landscape Approach in Action: Eight Years Later. In Reshaping Urban Conservation; Springer: Berlin/Heidelberg, Germany, 2019; pp. 21–54. [Google Scholar] [CrossRef]
- Bai, N.; Nourian, P.; Pereira Roders, A. Global Citizens and World Heritgae: Social Inclusion of Online Communities in Heritage Planning. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2021, XLVI-M-1-2021, 23–30. [Google Scholar] [CrossRef]
- UNESCO. Recommendation on the Historic Urban Landscape; Technical Report; UNESCO: Paris, France, 2011. [Google Scholar]
- Lu, W.; Stepchenkova, S. User-Generated Content as a Research Mode in Tourism and Hospitality Applications: Topics, Methods, and Software. J. Hosp. Mark. Manag. 2015, 24, 119–154. [Google Scholar] [CrossRef]
- Pickering, C.; Rossi, S.D.; Hernando, A.; Barros, A. Current Knowledge and Future Research Directions for the Monitoring and Management of Visitors in Recreational and Protected Areas. J. Outdoor Recreat. Tour. 2018, 21, 10–18. [Google Scholar] [CrossRef]
- Baltrusaitis, T.; Ahuja, C.; Morency, L.P. Multimodal Machine Learning: A Survey and Taxonomy. IEEE Trans. Pattern Anal. Mach. Intell. 2019, 41, 423–443. [Google Scholar] [CrossRef]
- Marine-Roig, E.; Anton Clavé, S. Tourism Analytics with Massive User-Generated Content: A Case Study of Barcelona. J. Destin. Mark. Manag. 2015, 4, 162–172. [Google Scholar] [CrossRef]
- Amato, F.; Cozzolino, G.; Di Martino, S.; Mazzeo, A.; Moscato, V.; Picariello, A.; Romano, S.; Sperlí, G. Opinions Analysis in Social Networks for Cultural Heritage Applications. Smart Innov. Syst. Technol. 2016, 55, 577–586. [Google Scholar] [CrossRef]
- Lee, H.; Kang, Y. Mining Tourists’ Destinations and Preferences through LSTM-based Text Classification and Spatial Clustering Using Flickr Data. Spat. Inf. Res. 2021, 29, 825–839. [Google Scholar] [CrossRef]
- Zhang, F.; Zhou, B.; Ratti, C.; Liu, Y. Discovering Place-informative Scenes and Objects Using Social Media photos. R. Soc. Open Sci. 2019, 6, 181375. [Google Scholar] [CrossRef]
- Giglio, S.; Bertacchini, F.; Bilotta, E.; Pantano, P. Using Social Media to Identify Tourism Attractiveness in Six Italian Cities. Tour. Manag. 2019, 72, 306–312. [Google Scholar] [CrossRef]
- Campillo-Alhama, C.; Martinez-Sala, A.M. Events 2.0 in the Transmedia Branding Strategy of World Cultural Heritage Sites. Prof. Inf. 2019, 28, e280509. [Google Scholar] [CrossRef]
- Liew, C.L. Participatory Cultural Heritage: A Tale of Two Institutions’ Use of Social Media. D-Lib Mag. 2014, 20, 1–17. [Google Scholar] [CrossRef]
- Williams, N.L.; Inversini, A.; Ferdinand, N.; Buhalis, D. Destination eWOM: A Macro and Meso Network Approach? Ann. Tour. Res. 2017, 64, 87–101. [Google Scholar] [CrossRef]
- Giglio, S.; Bertacchini, F.; Bilotta, E.; Pantano, P. Machine Learning and Point of Interests: Typical Tourist Italian Cities. Curr. Issues Tour. 2019, 23, 1646–1658. [Google Scholar] [CrossRef]
- Gabrielli, L.; Rinzivillo, S.; Ronzano, F.; Villatoro, D. From Tweets to Semantic Trajectories: Mining Anomalous Urban Mobility Patterns. In Citizen in Sensor Networks; Lecture Notes in Artificial Intelligence; Nin, J., Villatoro, D., Eds.; Springer: Berlin, Germany, 2014; Volume 8313, pp. 26–35. [Google Scholar] [CrossRef]
- Aggarwal, C.C. An Introduction to Social Network Data Analytics. In Social Network Data Analytics; Aggarwal, C.C., Ed.; Springer: Berlin/Heidelberg, Germany, 2011; Chapter 1; pp. 1–15. [Google Scholar] [CrossRef]
- Monteiro, V.; Henriques, R.; Painho, M.; Vaz, E. Sensing World Heritage An Exploratory Study of Twitter as a Tool for Assessing Reputation. In Proceedings of the Computational Science and Its Applications—ICCSA 2014, Guimarães, Portugal, 30 June–3 July 2014; Pt II. Volume 8580, pp. 404–419. [Google Scholar]
- Ginzarly, M.; Pereira Roders, A.; Teller, J. Mapping Historic Urban Landscape Values through Social Media. J. Cult. Herit. 2019, 36, 1–11. [Google Scholar] [CrossRef]
- Crandall, D.; Backstrom, L.; Huttenlocher, D.; Kleinberg, J. Mapping the World’s Photos. In Proceedings of the 18th International World Wide Web Conference—WWW’09, Madrid, Spain, 20–24 April 2009; pp. 761–770. [Google Scholar] [CrossRef]
- Gomez, R.; Gomez, L.; Gibert, J.; Karatzas, D. Learning from #Barcelona Instagram data what Locals and Tourists post about its Neighbourhoods. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops, Munich, Germany, 8–14 September 2018; pp. 530–544. [Google Scholar] [CrossRef]
- Kang, Y.; Cho, N.; Yoon, J.; Park, S.; Kim, J. Transfer Learning of a Deep Learning Model for Exploring Tourists’ Urban Image Using Geotagged Photos. ISPRS Int. J. Geo-Inf. 2021, 10, 137. [Google Scholar] [CrossRef]
- Cho, N.; Kang, Y.; Yoon, J.; Park, S.; Kim, J. Classifying Tourists’ Photos and Exploring Tourism Destination Image Using a Deep Learning Model. J. Qual. Assur. Hosp. Tour. 2022, 1–29. [Google Scholar] [CrossRef]
- Pan, S.J.; Yang, Q. A Survey on Transfer Learning. IEEE Trans. Knowl. Data Eng. 2009, 22, 1345–1359. [Google Scholar] [CrossRef]
- Tobler, W.R. A Computer Movie Simulating Urban Growth in the Detroit Region. Econ. Geogr. 1970, 46, 234–240. [Google Scholar] [CrossRef]
- Pearson, K. The Problem of the Random Walk. Nature 1905, 72, 294. [Google Scholar] [CrossRef]
- Batty, M. The New Science of Cities; MIT Press: Cambridge, MA, USA, 2013. [Google Scholar]
- Nourian, P. Configraphics: Graph Theoretical Methods for Design and Analysis of Spatial Configurations; TU Delft: Delft, The Netherlands, 2016. [Google Scholar]
- Ren, Y.; Cheng, T.; Zhang, Y. Deep Spatio-temporal Residual Neural Networks for Road-network-based Data Modeling. Int. J. Geogr. Inf. Sci. 2019, 33, 1894–1912. [Google Scholar] [CrossRef]
- Zhang, Y.; Cheng, T. Graph Deep Learning Model for Network-based Predictive Hotspot Mapping of Sparse Spatio-temporal Events. Comput. Environ. Urban Syst. 2020, 79, 101403. [Google Scholar] [CrossRef]
- Wasserman, S.; Faust, K. Social Network Analysis: Methods and Applications; Cambridge University Press: Cambridge, UK, 1994. [Google Scholar]
- Lazer, D.; Pentland, A.; Adamic, L.; Aral, S.; Barabasi, A.L.; Brewer, D.; Christakis, N.; Contractor, N.; Fowler, J.; Gutmann, M.; et al. Social science. Computational Social Science. Science 2009, 323, 721–723. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Barabási, A.L. Network science. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 2013, 371, 20120375. [Google Scholar] [CrossRef] [PubMed]
- Pentland, A. Social Physics: How Social Networks Can Make Us Smarter; Penguin: New York, NY, USA, 2015. [Google Scholar]
- Cheng, T.; Wicks, T. Event Detection Using Twitter: A Spatio-temporal Approach. PLoS ONE 2014, 9, e97807. [Google Scholar]
- Zhang, Z.; Cui, P.; Zhu, W. Deep Learning on Graphs: A Survey. IEEE Trans. Knowl. Data Eng. 2022, 34, 249–270. [Google Scholar] [CrossRef]
- Ma, Y.; Tang, J. Deep Learning on Graphs; Cambridge University Press: Cambridge, UK, 2021. [Google Scholar]
- Prince, M. Does Active Learning Work? A Review of the Research. J. Eng. Educ. 2004, 93, 223–231. [Google Scholar] [CrossRef]
- Nowak, S.; Rüger, S. How Reliable are Annotations via Crowdsourcing. In Proceedings of the International Conference on Multimedia Information Retrieval, Philadelphia, PA, USA, 29–31 March 2010; pp. 557–566. [Google Scholar] [CrossRef]
- Settles, B. From Theories to Queries: Active Learning in Practice. In Proceedings of the Active Learning and Experimental Design Workshop in Conjunction with AISTATS 2010, JMLR Workshop and Conference Proceedings, Sardinia, Italy, 16 May 2010; pp. 1–18. [Google Scholar]
- Zhou, Z.H.; Li, M. Semi-supervised Learning by Disagreement. Knowl. Inf. Syst. 2010, 24, 415–439. [Google Scholar] [CrossRef]
- Lee, D.H. Pseudo-label: The Simple and Efficient Semi-supervised Learning Method for Deep Neural Networks. In Proceedings of the Workshop on Challenges in Representation Learning, ICML 2013, Atlanta, GA, USA, 6–21 June 2013; Volume 3, p. 896. [Google Scholar]
- Kipf, T.N.; Welling, M. Semi-Supervised Classification with Graph Convolutional Networks. In Proceedings of the 5th International Conference on Learning Representations, ICLR 2017, Toulon, France, 24–26 April 2017; pp. 1–14. [Google Scholar]
- Pereira Roders, A. Revealing the World Heritage Cities and their Varied Natures. In Proceedings of the 2nd International Conference on Heritage and Sustainable Development (HERITAGE 2010), Evora, Portugal, 22–26 June 2010; pp. 245–253. [Google Scholar]
- Valese, M.; Noardo, F.; Pereira Roders, A. World Heritage Mapping in a Standard-based Structured Geographical Information System. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. 2020, XLIII-B4-2020, 81–88. [Google Scholar] [CrossRef]
- Jokilehto, J. Aesthetics in the World Heritage Context. In Proceedings of the International Conference of ICOMOS, ICCROM and Fondazione Romualdo Del Bianco, Florence, Italy, 2–4 March 2007; Polistampa: Florence, Italy, 2008; pp. 183–194. [Google Scholar]
- Jokilehto, J. What Is OUV? Defining the Outstanding Universal Value of Cultural World Heritage Properties; Technical Report; ICOMOS: Berlin, Germany, 2008. [Google Scholar]
- Bai, N.; Nourian, P.; Luo, R.; Pereira Roders, A. “What is OUV” Revisited: A Computational Interpretation on the Statements of Outstanding Universal Value. ISPRS Ann. Photogramm. Remote. Sens. Spat. Inf. Sci. 2021, VIII-M-1-2021, 25–32. [Google Scholar] [CrossRef]
- Van Dijck, J. Flickr and the Culture of Connectivity: Sharing views, Experiences, Memories. Mem. Stud. 2011, 4, 401–415. [Google Scholar] [CrossRef]
- Lin, T.Y.; Maire, M.; Belongie, S.; Hays, J.; Perona, P.; Ramanan, D.; Dollár, P.; Zitnick, C.L. Microsoft COCO: Common objects in context. In Proceedings of the European Conference on Computer Vision, Zurich, Switzerland, 6–12 September 2014; Springer: Berlin/Heidelberg, Germany, 2014; pp. 740–755. [Google Scholar]
- Tenkanen, H.; Di Minin, E.; Heikinheimo, V.; Hausmann, A.; Herbst, M.; Kajala, L.; Toivonen, T. Instagram, Flickr, or Twitter: Assessing the Usability of Social Media Data for Visitor Monitoring in Protected Areas. Sci. Rep. 2017, 7, 17615. [Google Scholar] [CrossRef] [PubMed]
- Li, D.; Zhou, X.; Wang, M. Analyzing and Visualizing the Spatial Interactions between Tourists and Locals: A Flickr Study in ten US Cities. Cities 2018, 74, 249–258. [Google Scholar] [CrossRef]
- Huiskes, M.J.; Lew, M.S. The MIR Flickr Retrieval Evaluation. In Proceedings of the 1st International ACM Conference on Multimedia Information Retrieval, MIR2008, Co-Located with the 2008 ACM International Conference on Multimedia, MM’08, Vancouver, BC, Canada, 30–31 October 2008; pp. 39–43. [Google Scholar] [CrossRef]
- Chua, T.S.; Tang, J.; Hong, R.; Li, H.; Luo, Z.; Zheng, Y. NUS-WIDE: A Real-world Web Image Database from National University of Singapore. In Proceedings of the ACM International Conference on Image and Video Retrieval, Santorini Island, Greece, 8–10 July 2009; pp. 1–9. [Google Scholar]
- Tang, L.; Liu, H. Relational Learning via Latent Social Dimensions. In Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Paris, France, 28 June–1 July 2009; pp. 817–826. [Google Scholar]
- Deng, J.; Dong, W.; Socher, R.; Li, L.J.; Li, K.; Li, F.-F. ImageNet: A Large-scale Hierarchical Image Database. In Proceedings of the 2009 IEEE Conference on Computer Vision and Pattern Recognition, Miami, FL, USA, 20–25 June 2009; IEEE: New York, NY, USA, 2009; pp. 248–255. [Google Scholar]
- Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet Classification with Deep Convolutional Neural Networks. In Proceedings of the Annual Conference on Neural Information Processing Systems, Lake Tahoe, CA, USA, 3–8 December 2012. [Google Scholar]
- Plummer, B.A.; Wang, L.; Cervantes, C.M.; Caicedo, J.C.; Hockenmaier, J.; Lazebnik, S. Flickr30k Entities: Collecting Region-to-phrase Correspondences for Richer Image-to-sentence Models. In Proceedings of the IEEE International Conference on Computer Vision, Santiago, Chile, 7–13 December 2015; pp. 2641–2649. [Google Scholar]
- Zhou, Y.; Long, Y. SinoGrids: A Practice for Open Urban Data in China. Cartogr. Geogr. Inf. Sci. 2016, 43, 379–392. [Google Scholar] [CrossRef]
- Zeng, H.; Zhou, H.; Srivastava, A.; Kannan, R.; Prasanna, V. GraphSAINT: Graph Sampling based Inductive Learning Method. arXiv 2019, arXiv:1907.04931. [Google Scholar]
- Zhou, B.; Lapedriza, A.; Xiao, J.; Torralba, A.; Oliva, A. Learning Deep Features for Scene Recognition Using Places Database. In Proceedings of the Annual Conference on Neural Information Processing Systems, Montreal, QC, Canada, 8–13 December 2014; pp. 1–9. [Google Scholar]
- Zhou, B.; Lapedriza, A.; Khosla, A.; Oliva, A.; Torralba, A. Places: A 10 Million Image Database for Scene Recognition. IEEE Trans. Pattern Anal. Mach. Intell. 2017, 40, 1452–1464. [Google Scholar] [CrossRef]
- Patterson, G.; Hays, J. SUN Attribute Database: Discovering, Annotating, and Recognizing Scene Attributes. In Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition, Providence, RI, USA, 16–21 June 2012; IEEE: New York, NY, USA, 2012; pp. 2751–2758. [Google Scholar]
- Patterson, G.; Xu, C.; Su, H.; Hays, J. The SUN Attribute Database: Beyond Categories for Deeper Scene Understanding. Int. J. Comput. Vis. 2014, 108, 59–81. [Google Scholar] [CrossRef]
- He, K.; Zhang, X.; Ren, S.; Sun, J. Deep Residual Learning for Image Recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA, 27–30 June 2016; pp. 770–778. [Google Scholar]
- Schroff, F.; Kalenichenko, D.; Philbin, J. FaceNet: A Unified Embedding for Face Recognition and Clustering. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA, 7–12 June 2015; pp. 815–823. [Google Scholar]
- Cao, Q.; Shen, L.; Xie, W.; Parkhi, O.M.; Zisserman, A. VGGFace2: A Dataset for Recognising Faces across Pose and Age. In Proceedings of the 2018 13th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2018), Xi’an, China, 15–19 May 2018; IEEE: New York, NY, USA, 2018; pp. 67–74. [Google Scholar]
- Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A.N.; Kaiser, L.; Polosukhin, I. Attention is All you Need. In In Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, CA, USA, 4–9 December 2017; pp. 6000–6010. [Google Scholar]
- Devlin, J.; Chang, M.W.; Lee, K.; Toutanova, K. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Minneapolis, MN, USA, 2–7 June 2019; Association for Computational Linguistics: Stroudsburg, PA, USA, 2019; Volume 1 (Long and Short Papers), pp. 4171–4186. [Google Scholar]
- Clark, K.; Khandelwal, U.; Levy, O.; Manning, C.D. What Does BERT Look at? In An Analysis of BERT’s Attention. In Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, BlackboxNLP@ACL 2019, Florence, Italy, 1 August 2019; Association for Computational Linguistics: Stroudsburg, PA, USA, 2019; pp. 276–286. [Google Scholar] [CrossRef]
- Sun, C.; Qiu, X.; Xu, Y.; Huang, X. How to Fine-Tune BERT for Text Classification? In Proceedings of the China National Conference on Chinese Computational Linguistics, Kunming, China, 18–20 October 2019; Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer: Cham, Switzerland, 2019; Volume 11856 LNAI, pp. 194–206. [Google Scholar] [CrossRef]
- Bai, N.; Luo, R.; Nourian, P.; Pereira Roders, A. WHOSe Heritage: Classification of UNESCO World Heritage Statements of ”Outstanding Universal Value” with Soft Labels. In Findings of the Association for Computational Linguistics: EMNLP 2021; Association for Computational Linguistics: Punta Cana, Dominican Republic, 2021; pp. 366–384. [Google Scholar]
- Reiter, R. Towards a Logical Reconstruction of Relational Database Theory. In Topics in Information Systems; Brodie, M.L., Mylopoulos, J., Schmidt, J.W., Eds.; Springer: New York, NY, USA, 1984; pp. 191–238. [Google Scholar] [CrossRef]
- Nourian, P.; Rezvani, S.; Sariyildiz, I.; van der Hoeven, F. Spectral Modelling for Spatial Network Analysis. In Proceedings of the Symposium on Simulation for Architecture and Urban Design (simAUD 2016), London, UK, 16–18 May 2016; pp. 103–110. [Google Scholar]
- Hillier, B.; Hanson, J. The Social Logic of Space; Cambridge University Press: Cambridge, UK, 1989. [Google Scholar]
- Penn, A. Space Syntax and Spatial Cognition: Or why the Axial Line? Environ. Behav. 2003, 35, 30–65. [Google Scholar] [CrossRef]
- Ratti, C. Space Syntax: Some Inconsistencies. Environ. Plan. Plan. Des. 2004, 31, 487–499. [Google Scholar] [CrossRef]
- Blanchard, P.; Volchenkov, D. Mathematical Analysis of Urban Spatial Networks; Springer: Berlin, Germany, 2008. [Google Scholar]
- Boeing, G. OSMnx: New Methods for Acquiring, Constructing, Analyzing, and Visualizing Complex Street Networks. Comput. Environ. Urban Syst. 2017, 65, 126–139. [Google Scholar] [CrossRef] [Green Version]
- Howley, P.; Scott, M.; Redmond, D. Sustainability versus Liveability: An Investigation of Neighbourhood Satisfaction. J. Environ. Plan. Manag. 2009, 52, 847–864. [Google Scholar] [CrossRef]
- Howard, J.; Ruder, S. Universal Language Model Fine-tuning for Text Classification. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018, Melbourne, Australia, 15–20 July 2018; Association for Computational Linguistics: Stroudsburg, PA, USA, 2018; Volume 1: Long Papers, pp. 328–339. [Google Scholar] [CrossRef]
- Zhu, X.; Goldberg, A.B. Introduction to Semi-supervised Learning. Synth. Lect. Artif. Intell. Mach. Learn. 2009, 3, 1–130. [Google Scholar]
- Pustejovsky, J.; Stubbs, A. Natural Language Annotation for Machine Learning: A Guide to Corpus-Building for Applications; O’Reilly Media, Inc.: Sebastopol, CA, USA, 2012. [Google Scholar]
- Sohn, K.; Berthelot, D.; Carlini, N.; Zhang, Z.; Zhang, H.; Raffel, C.A.; Cubuk, E.D.; Kurakin, A.; Li, C.L. FixMatch: Simplifying Semi-supervised Learning with Consistency and Confidence. Adv. Neural Inf. Process. Syst. 2020, 33, 596–608. [Google Scholar]
- Veldpaus, L.; Roders, A.P. Learning from a Legacy: Venice to Valletta. Chang. Over Time 2014, 4, 244–263. [Google Scholar] [CrossRef]
- Gustcoven, E. Attributes of World Heritage Cities, Sustainability by Management—A Comparative Study between the World Heritage Cities of Amsterdam, Edinburgh and Querétaro. Master’s Thesis, KU Leuven, Leuven, Belgium, 2016. [Google Scholar]
- UNESCO. Heritage in Urban Contexts: Impact of Development Projects on World Heritage Properties in Cities; Technical Report; UNESCO World Heritage Centre: Paris, France, 2020. [Google Scholar]
- Pedregosa, F.; Varoquaux, G.; Gramfort, A.; Michel, V.; Thirion, B.; Grisel, O.; Blondel, M.; Prettenhofer, P.; Weiss, R.; Dubourg, V.; et al. Scikit-learn: Machine Learning in Python. J. Mach. Learn. Res. 2011, 12, 2825–2830. [Google Scholar]
- Hinton, G.E. Connectionist Learning Procedures. In Machine Learning; Elsevier: Amsterdam, The Netherlands, 1990; pp. 555–610. [Google Scholar]
- Altman, N.S. An Introduction to Kernel and Nearest-neighbor Nonparametric Regression. Am. Stat. 1992, 46, 175–185. [Google Scholar]
- Rish, I. An Empirical Study of the Naive Bayes Classifier. In Proceedings of the IJCAI 2001 Workshop on Empirical Methods in Artificial Intelligence, Seattle, WA, USA, 4–10 August 2001; Volume 3, pp. 41–46. [Google Scholar]
- Platt, J. Probabilistic Outputs for Support Vector Machines and Comparisons to Regularized Likelihood Methods. Adv. Large Margin Classif. 1999, 10, 61–74. [Google Scholar]
- Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef]
- Breiman, L. Bagging Predictors. Mach. Learn. 1996, 24, 123–140. [Google Scholar] [CrossRef]
- Arlot, S.; Celisse, A. A Survey of Cross-validation Procedures for Model Selection. Stat. Surv. 2010, 4, 40–79. [Google Scholar] [CrossRef]
- Zhou, Z.H. Ensemble Methods: Foundations and Algorithms; CRC Press: Boca Raton, FL, USA, 2012. [Google Scholar]
- Breiman, L. Stacked Regressions. Mach. Learn. 1996, 24, 49–64. [Google Scholar] [CrossRef]
- Hagberg, A.; Swart, P.; S Chult, D. Exploring Network Structure, Dynamics, and Function Using NetworkX; Technical Report; Los Alamos National Lab. (LANL): Los Alamos, NM, USA, 2008. [Google Scholar]
- Karypis, G.; Kumar, V. Analysis of Multilevel Graph Partitioning. In Proceedings of the 1995 ACM/IEEE Conference on Supercomputing—Supercomputing’95, San Diego, CA, USA, 3–8 December 1995; IEEE: New York, NY, USA, 1995; p. 29. [Google Scholar] [CrossRef]
- Lafon, S.; Lee, A.B. Diffusion Maps and Coarse-Graining: A Unified Framework for Dimensionality Reduction, Graph Partitioning, and Data Set Parameterization. IEEE Trans. Pattern Anal. Mach. Intell. 2006, 28, 1393–1403. [Google Scholar] [CrossRef]
- Gao, H.; Ji, S. Graph U-Nets. In Proceedings of the International Conference on Machine Learning, PMLR 2019, Long Beach, CA, USA, 9–15 June 2019; pp. 2083–2092. [Google Scholar]
- Encalada-Abarca, L.; Ferreira, C.C.; Rocha, J. Measuring Tourism Intensification in Urban Destinations: An Approach based on Fractal Analysis. J. Travel Res. 2022, 61, 394–413. [Google Scholar] [CrossRef]
- Eom, Y.H.; Jo, H.H. Tail-scope: Using Friends to Estimate Heavy Tails of Degree Distributions in Large-scale Complex Networks. Sci. Rep. 2015, 5, 09752. [Google Scholar] [CrossRef]
- Psarra, S. The Venice Variations: Tracing the Architectural Imagination; UCL Press: London, UK, 2018. [Google Scholar]
- Blum, A.; Mitchell, T. Combining Labeled and Unlabeled Data with Co-training. In Proceedings of the Eleventh Annual Conference on Computational Learning Theory, Madison, WI, USA, 24–26 July 1998; pp. 92–100. [Google Scholar]
- Wang, Y.; Yao, Q.; Kwok, J.T.; Ni, L.M. Generalizing from a few Examples: A Survey on Few-shot Learning. ACM Comput. Surv. 2020, 53, 1–34. [Google Scholar] [CrossRef]
- Veličković, P.; Cucurull, G.; Casanova, A.; Romero, A.; Lió, P.; Bengio, Y. Graph Attention Networks. In Proceedings of the 6th International Conference on Learning Representations, ICLR 2018, Vancouver, BC, Canada, 30 April–3 May 2018; pp. 1–12. [Google Scholar]
- Miah, S.J.; Vu, H.Q.; Gammack, J.; McGrath, M. A Big Data Analytics Method for Tourist Behaviour Analysis. Inf. Manag. 2017, 54, 771–785. [Google Scholar] [CrossRef]
- Adamic, L.A.; Adar, E. Friends and Neighbors on the Web. Soc. Netw. 2003, 25, 211–230. [Google Scholar] [CrossRef]
- Majid, A.; Chen, L.; Chen, G.; Mirza, H.T.; Hussain, I.; Woodward, J. A Context-aware Personalized Travel Recommendation System based on Geotagged Social Media Data Mining. Int. J. Geogr. Inf. Sci. 2013, 27, 662–684. [Google Scholar] [CrossRef]
- Pang, Y.; Zhao, Y.; Li, D. Graph Pooling via Coarsened Graph Infomax. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval, Virtual, 11–15 July 2021; pp. 2177–2181. [Google Scholar]
- Zhang, M.; Cui, Z.; Neumann, M.; Chen, Y. An End-to-end Deep Learning Architecture for Graph Classification. In Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence, New Orleans, LA, USA, 2–7 February 2018; pp. 4438–4445. [Google Scholar]
- Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative Adversarial Nets. In Proceedings of the Annual Conference on Neural Information Processing Systems, Montreal, QC, Canada, 8–13 December 2014; pp. 2672–2680. [Google Scholar]
- Ma, Y.; Ren, Z.; Jiang, Z.; Tang, J.; Yin, D. Multi-dimensional Network Embedding with Hierarchical Structure. In Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining, Los Angeles, CA, USA, 5–9 February 2018; pp. 387–395. [Google Scholar]
- Nguyen, G.H.; Lee, J.B.; Rossi, R.A.; Ahmed, N.K.; Koh, E.; Kim, S. Continuous-time Dynamic Network Embeddings. In Proceedings of the Companion Proceedings of the The Web Conference 2018, Lyon, France, 23–27 April 2018; pp. 969–976. [Google Scholar]
- Yuster, R.; Zwick, U. Fast Sparse Matrix Multiplication. ACM Trans. Algorithms 2005, 1, 2–13. [Google Scholar] [CrossRef] [Green Version]
- Urry, J.; Larsen, J. The Tourist Gaze 3.0; SAGE: Thousand Oaks, CA, USA, 2011. [Google Scholar]
- Bekker, R. Creating Insights in Tourism with Flickr Photography, Visualizing and Analysing Spatial and Temporal Patterns in Venice. Master’s Thesis, Rijksuniversiteit Groningen, Groningen, The Netherlands, 2020. [Google Scholar]
City | Geo-Location | WHL Name | OUV Criteria | Area of Property | Inscription Date |
---|---|---|---|---|---|
Amsterdam (AMS) | 52.365000N 4.887778E | Seventeenth-Century Canal Ring Area of Amsterdam inside the Singelgracht | (i), (ii), (iv) | 198.2 ha | 2010 |
Suzhou (SUZ) | 31.302300N 120.631300E | Classical Gardens of Suzhou | (i), (ii), (iii), (iv), (v) | 11.9 ha | 2000 |
Venice (VEN) | 45.438759N 12.327145E | Venice and its Lagoon | (i), (ii), (iii), (iv), (v), (vi) | 70,176.2 ha | 1987 |
City | AMS | SUZ | VEN | VEN-XL |
---|---|---|---|---|
IDs Collected | 5000 | 4229 | 5000 | 116,675 |
Is Downloadable | 3727 | 3137 | 2952 | 80,964 |
Downloaded Posts | 3727 | 3137 | 2951 | 80,963 |
Has Textual Data * | 3404 | 2692 | 2801 | 77,644 |
Has Unique Texts ** | 3130 | 1963 | 1952 | 59,396 |
Unique Sentences | 2247 | 361 | 3249 | 61,253 |
Original Posts ** | 2904 | 754 | 1761 | 49,823 |
Posting Owners | 195 | 95 | 330 | 6077 |
Sets to Calculate IoU Jaccard Index | AMS | SUZ | VEN |
---|---|---|---|
# Compared Posts w. Visual Features | 3727 | 3137 | 2951 |
Top-1 scene predictions | 0.656 | 0.676 | 0.704 |
— | (0.475) | (0.468) | (0.456) |
Top-5 scene predictions | 0.615 | 0.636 | 0.635 |
— | (0.179) | (0.238) | (0.229) |
Top-1 attribute predictions | 0.867 | 0.853 | 0.838 |
— | (0.339) | (0.354) | (0.368) |
Top-10 attribute predictions | 0.820 | 0.802 | 0.819 |
— | (0.140) | (0.144) | (0.139) |
# Compared Posts w. Textual Features | 2904 | 754 | 1761 |
Top-1 OUV predictions | 0.775 | 0.923 | 0.714 |
— | (0.418) | (0.267) | (0.452) |
Top-3 OUV predictions | 0.840 | 0.938 | 0.791 |
— | (0.246) | (0.182) | (0.266) |
Features | AMS | SUZ | VEN | VEN-XL |
---|---|---|---|---|
# Posts w. Faces | 667 | 303 | 166 | 9287 |
# Faces detected | 1.547 | 1.403 | 1.349 | 1.298 |
— | (0.830) | (0.707) | (0.785) | (0.651) |
Model Confidence | 0.955 | 0.956 | 0.930 | 0.948 |
— | (0.079) | (0.081) | (0.099) | (0.081) |
Area proportion of faces | 0.049 | 0.057 | 0.077 | 0.076 |
— | (0.112) | (0.073) | (0.185) | (0.112) |
# Posts w. Texts * | 2904 | 754 | 1761 | 49,823 |
# Posts in English | 1488 | 368 | 640 | 20,271 |
# Posts in Native Lang | 1773 | 27 | 1215 | 28,633 |
# Posts in Other Lang | 536 | 413 | 657 | 21,916 |
ML Model | CV Acc | Val Acc | Val F1 | Test Acc | Test F1 |
---|---|---|---|---|---|
MLP | 0.767 | 0.749 | 0.70 | 0.789 | 0.72 |
KNN | 0.756 | 0.724 | 0.67 | 0.767 | 0.71 |
GNB | 0.738 | 0.749 | 0.71 | 0.800 | 0.77 |
SVM | 0.797 | 0.754 | 0.71 | 0.822 | 0.78 |
RF | 0.766 | 0.734 | 0.68 | 0.789 | 0.72 |
BC-SVM | 0.780 | 0.759 | 0.71 | 0.811 | 0.74 |
VOTE | 0.788 | 0.764 | 0.72 | 0.855 | 0.82 |
STACK | 0.794 | 0.768 | 0.72 | 0.844 | 0.81 |
Graph Features | AMS | SUZ | VEN | VEN-XL |
---|---|---|---|---|
# Nodes in V | 788 | 230 | 915 | 3549 |
# Edges in E | 3331 | 680 | 10,385 | 120,033 |
# Connected Components | 72 | 38 | 6 | 13 |
# Nodes Largest CC * | 355 | 50 | 897 | 3498 |
Graph Density | 0.011 | 0.026 | 0.025 | 0.019 |
# Isolated Nodes in | 157 | 88 | 20 | 22 |
Graph Features | AMS | SUZ | VEN |
---|---|---|---|
Temporal Graph | |||
# Nodes * | 3727 | 3137 | 2951 |
# Edges | 692,839 | 293,328 | 249,120 |
Diameter | 145 | 116 | 270 |
Graph Density | 0.100 | 0.060 | 0.057 |
Social Graph | |||
# Nodes ** | 3696 | 3120 | 2916 |
# Edges | 877,584 | 602,821 | 242,576 |
# Connected Components | 47 | 56 | 60 |
# Nodes Largest CC | 2694 | 942 | 2309 |
Diameter Largest CC | 7 | 6 | 10 |
Graph Density | 0.129 | 0.124 | 0.057 |
Spatial Graph | |||
# Nodes ** | 3632 | 3102 | 2938 |
# Edges | 135,079 | 415,049 | 221,414 |
# Connected Components | 134 | 91 | 13 |
# Nodes Largest CC | 1485 | 829 | 2309 |
Diameter Largest CC | 22 | 1 | 22 |
Graph Density | 0.020 | 0.086 | 0.051 |
Simple Composed Graph | |||
# Nodes * | 3727 | 3137 | 2951 |
# Edges | 1,271,171 | 916,496 | 534,513 |
Diameter | 4 | 5 | 4 |
Graph Density | 0.183 | 0.186 | 0.123 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Bai, N.; Nourian, P.; Luo, R.; Pereira Roders, A. Heri-Graphs: A Dataset Creation Framework for Multi-Modal Machine Learning on Graphs of Heritage Values and Attributes with Social Media. ISPRS Int. J. Geo-Inf. 2022, 11, 469. https://doi.org/10.3390/ijgi11090469
Bai N, Nourian P, Luo R, Pereira Roders A. Heri-Graphs: A Dataset Creation Framework for Multi-Modal Machine Learning on Graphs of Heritage Values and Attributes with Social Media. ISPRS International Journal of Geo-Information. 2022; 11(9):469. https://doi.org/10.3390/ijgi11090469
Chicago/Turabian StyleBai, Nan, Pirouz Nourian, Renqian Luo, and Ana Pereira Roders. 2022. "Heri-Graphs: A Dataset Creation Framework for Multi-Modal Machine Learning on Graphs of Heritage Values and Attributes with Social Media" ISPRS International Journal of Geo-Information 11, no. 9: 469. https://doi.org/10.3390/ijgi11090469
APA StyleBai, N., Nourian, P., Luo, R., & Pereira Roders, A. (2022). Heri-Graphs: A Dataset Creation Framework for Multi-Modal Machine Learning on Graphs of Heritage Values and Attributes with Social Media. ISPRS International Journal of Geo-Information, 11(9), 469. https://doi.org/10.3390/ijgi11090469