An intelligent agent approach for visual information structure generation

Duman, Hakan; Healing, Alex; Ghanea-Hercock, Robert

doi:10.1109/ia.2009.4927500

“…Systems presented for clustering and exploration in Duman, Healing & Ghanea-Hercock (2009), Desjardins, MacGlashan & Ferraioli (2007, Beale (2007) and Lee et al (2012) transform the data into a spring-embedded graph layout, encoding the distance between the elements as forces in the force-directed layout. The objective in these systems is the projection of the distances in a reduced dimension allowing clustering assignment using partitioning-based methods.…”

Section: Graph Representationmentioning

confidence: 99%

MCLEAN: Multilevel Clustering Exploration As Network

Alcaide

¹

,

Aerts

²

2018

PeerJ Computer Science

View full text Add to dashboard Cite

Finding useful patterns in datasets has attracted considerable interest in the field of visual analytics. One of the most common tasks is the identification and representation of clusters. However, this is non-trivial in heterogeneous datasets since the data needs to be analyzed from different perspectives. Indeed, highly variable patterns may mask underlying trends in the dataset. Dendrograms are graphical representations resulting from agglomerative hierarchical clustering and provide a framework for viewing the clustering at different levels of detail. However, dendrograms become cluttered when the dataset gets large, and the single cut of the dendrogram to demarcate different clusters can be insufficient in heterogeneous datasets. In this work, we propose a visual analytics methodology called MCLEAN that offers a general approach for guiding the user through the exploration and detection of clusters. Powered by a graphbased transformation of the relational data, it supports a scalable environment for representation of heterogeneous datasets by changing the spatialization. We thereby combine multilevel representations of the clustered dataset with community finding algorithms. Our approach entails displaying the results of the heuristics to users, providing a setting from which to start the exploration and data analysis. To evaluate our proposed approach, we conduct a qualitative user study, where participants are asked to explore a heterogeneous dataset, comparing the results obtained by MCLEAN with the dendrogram. These qualitative results reveal that MCLEAN is an effective way of aiding users in the detection of clusters in heterogeneous datasets. The proposed methodology is implemented in an R package available at https://bitbucket.org/vda-lab/mclean.

show abstract

“…Systems presented for clustering and exploration in Duman et al (2009), Desjardins et al (2007), Beale (2007) and, Lee et al (2012) transform the data into a spring-embedded graph layout, encoding the distance between the elements as forces in the force-directed layout. The objective in these systems is the The network exploration of MCLEAN can be considered close to the solutions proposed for the navigation of the clustering results for large-scale graph visualization systems, such as Eades and Feng (1996) and Eades and Huang (2000).…”

Section: Graph Representationmentioning

confidence: 99%

MCLEAN: Multilevel Clustering Exploration As Network

Alcaide

¹

,

Aerts

²

2017

Preprint

0

View full text Add to dashboard Cite

Finding useful patterns in datasets has attracted considerable interest in the field of visual analytics. One of the most common tasks is the identification and representation of clusters. However, this is non-trivial in heterogeneous datasets since the data needs to be analyzed from different perspectives. Indeed, highly variable patterns may mask underlying trends in the dataset. Dendrograms are graphical representations resulting from agglomerative hierarchical clustering and provide a framework for viewing the clustering at different levels of detail. However, dendrograms become cluttered when the dataset gets large, and the single cut of the dendrogram to demarcate different clusters can be insufficient in heterogeneous datasets.In this work, we propose a visual analytics methodology called MCLEAN that offers a general approach for guiding the user through the exploration and detection of clusters.Powered by a graph-based transformation of the relational data, it supports a scalable environment for representation of heterogeneous datasets by changing the spatialization.We thereby combine multilevel representations of the clustered dataset with community finding algorithms. Our approach entails displaying the results of the heuristics to users, providing a setting from which to start the exploration and data analysis.To evaluate our proposed approach, we conduct a qualitative user study, where Finding useful patterns in datasets has attracted considerable interest in the field of visual analytics. One of the most common tasks is the identification and representation of clusters. However, this is non-trivial in heterogeneous datasets since the data needs to be analyzed from different perspectives. Indeed, highly variable patterns may mask underlying trends in the dataset. Dendrograms are graphical representations resulting from agglomerative hierarchical clustering and provide a framework for viewing the clustering at different levels of detail. However, dendrograms become cluttered when the dataset gets large, and the single cut of the dendrogram to demarcate different clusters can be insufficient in heterogeneous datasets. In this work, we propose a visual analytics methodology called MCLEAN that offers a general approach for guiding the user through the exploration and detection of clusters. Powered by a graph-based transformation of the relational data, it supports a scalable environment for representation of heterogeneous datasets by changing the spatialization. We thereby combine multilevel representations of the clustered dataset with community finding algorithms. Our approach entails displaying the results of the heuristics to users, providing a setting from which to start the exploration and data analysis. To evaluate our proposed approach, we conduct a qualitative user study, where participants are asked to explore a heterogeneous dataset, comparing the results obtained by MCLEAN with the dendrogram. These qualitative results reveal that MCLEAN is an effective way of aiding users in the detection of...

show abstract

“…Cyclone, as described in our previous works [11], [15], is an intelligent agent-based visual environment proving a means for the user to exploit, analyze and structure unstructured information into a more manageable form. The main strength of Cyclone is its capability to couple data mining techniques with intuitive information visualization and an adaptive learning system, creating a feedback loop between the user and the system.…”

Section: B Cyclonementioning

confidence: 99%

“…Furthermore, the communities are extended to incorporate other pertinent entities (such as keywords, posts and users) so as to widen the coverage of communities in the forum. The formed overlapping communities provide also the foundation for Cyclone [11], which is a visual environment allowing the user to explore, analyze, fine-tune the extracted communities simply by altering (e.g. adding new keywords, removing or amending) the set of keywords associated with the communities.…”

Section: Introductionmentioning

confidence: 99%

Overlapping Communities Generation for Online Support Forums

Wang

¹

,

Duman

²

,

Nguyen

³

et al. 2009

2009 International Conference on Adaptive and Intelligent Systems

Self Cite

View full text Add to dashboard Cite

This paper presents a novel approach of overlapping communities generation for online support forums. Different from traditional online forums that provide the most similar or relevant information to respond to a user query, the approach proposed in this paper manages online forums and provides user support based on overlapping communities. Inspired by natural societies, a forum is deemed as a complex network in which all entities (keywords, posts and user) of an online forum are grouped into a series of communities that can share members with each other. To enable this, a kind of keyword association graph is constructed based on the co-occurrences of keywords in user posts. CPM (Clique Percolation Method) is then applied to discover closely connected cliques (core clusters) in the graph. The core keyword clusters absorb related posts and users to form communities and the communities are naturally overlapping. The communities are also extended to include other un-clustered but relevant posts and users so all entities in the forum belong to at least one community. Overlapping communities in online forums provide a useful means to support various services including recommendation, alerting and profiling customer support.

show abstract

An intelligent agent approach for visual information structure generation

Cited by 5 publications

References 17 publications

MCLEAN: Multilevel Clustering Exploration As Network

MCLEAN: Multilevel Clustering Exploration As Network

MCLEAN: Multilevel Clustering Exploration As Network

Overlapping Communities Generation for Online Support Forums

Contact Info

Product

Resources

About