Fig 2 - uploaded by Tobias Kötter
Content may be subject to copyright.
Network and the corresponding adjacency list which serves as the transaction database for the frequent item set mining algorithm
Source publication
This article proposes a new approach to extract existing (or detect missing) concepts from a loosely integrated collection of information units by means of concept graph detection. Thereby a concept graph defines a concept by a quasi bipartite sub-graph of a bigger network with the members of the concept as the first vertex partition and their shar...
Context in source publication
Context 1
... order to apply frequent item set mining algorithms to find concept graphs in BisoNets we use the adjacency list of the network as transaction database. Therefore, for each vertex in the BisoNet, we create an entry in the transaction database with the vertex as the identifier and its direct neighbors as the products Figure 2). Once the database has been created we can apply frequent item set mining algorithms to detect vertices that share some neighbors. ...
Similar publications
For a vertex v of a connected graph G(V, E) with vertex set V(G), edge set E(G) and S ⊆ V(G). Given an ordered partition Π = {S1, S2, S3, …, Sk} of the vertex set V of G, the representation of a vertex v ∈ V with respect to Π is the vector r(v|Π) = (d(v, S1), d(v, S2), …, d(v, Sk)), where d(v, Sk) represents the distance between the vertex v and th...
This study proposes a system to estimate the depth order of regions belonging to a monocular image sequence. For each frame, the regions are ordered according to their relative depth using information from the previous and following frames. The algorithm estimates occlusions relying on a hierarchical region-based representation of the image by mean...
For a vertex v of a connected graph G and a subset S of V (G), the distance between v and S, denoted by d(v, S), is min{d(v, x) | x ∈ S}. Let Π = {S1, S2 . Sκ} be an ordered κ-partition of V (G). The representation of v with respect to Π is the κ-vector r(v|Π) = (d(v, S1), d(v, S2) . d(v, Sκ)). The κ-partition is a resolving partition if the k-vect...
![Figure 4. Graph F 4 [n]](profile/Rashid-Farooq/publication/326574158/figure/fig1/AS:652515838271488@1532583307424/Graph-F-4-n_Q320.jpg)
Let $G=(V(G),E(G))$ be a connected graph and $\Pi=\{S_{1},S_2,\dots,S_{k}\}$ be a $k$-partition of $V(G)$.The representation $r(v|\Pi)$ of a vertex $v$ with respect to $\Pi$ is the vector $(d(v,S_{1}),d(v,S_2),\dots,d(v,S_{k}))$, where $d(v,S_{i})=\min\{d(v,s_{i})\mid s_{i}\in S_{i}\}$.The partition $\Pi$ is called a resolving partition of $G$ if $...
The concept of a graph partition dimension was introduced by Chartrand et al. (1998). Let Π = {L1, L2, L3, · · ·, Lk } be a k-partition of V(G). The representation r(v|Π) of a vertex v with respect to Π is the vector (d(v, L1), d(v, L2), · · ·, d(v, Lk)). The partition Π is called a resolving partition of G if r(w|Π) ≠ r(v|Π) for all distinct w, v...
Citations
... Given an information network (Fig. 2), we want to identify underlying concepts (Fig. 3) and analyze them. In order to do so we extract concept graphs [10] that allow the discovery and description of concepts in such networks. Concept graphs allow the organization of information by grouping together objects (the members) that show common properties (the aspects), improving understanding of the concept graph and the actual concept it represents. ...
... Novelty of the proposed method over competitors: The existing work [10] on identifying concept graphs is restricted to identifying only perfect concept graphs i.e. concept graphs where all members share all properties. To overcome this drawback we propose a new fault-tolerant algorithm called FCDA (Fault-tolerant Concept Detection Algorithm) to find imperfect concept graphs as well. ...
... The general idea of concept graphs is to find disjoint sets of vertices V M , V A ⊆ V such that each vertex of the member set V M is connected to many vertices in the aspect set V A , and vice versa. The existing method [10], which is based on frequent itemsets, was defined to identify perfect concept graphs. A perfect concept graph is one that forms a quasi biclique where each member is connected to all aspects of the concept. ...
Given information about medical drugs and their properties,
how can we automatically discover that Aspirin
has blood-thinning properties, and thus prevents heart attacks?
Expressed in more general terms, if we have a large information network that integrates data from heterogeneous data sources, how can we extract semantic information that provides a better understanding of the integrated data and also helps us to identify missing links?
We propose to extract concepts that describe groups of objects and their common properties from the integrated data.
The discovered concepts provide semantic information as well as an abstract view on the integrated data and thus improve the understanding of complex systems.
Our proposed method has the following desirable properties:
(a) it is parameter-free and therefore requires no user-defined parameters
(b) it is fault-tolerant, allowing for the detection of missing links and
(c) it is scalable, being linear on the input size.
We demonstrate the effectiveness and scalability of the proposed method
on real, publicly available graphs.
... Metaphors play a major role in our everyday life as they afford a degree of flexibility that facilitates discoveries by connecting seemingly unrelated subjects [39]. A first approach to detect bridging concepts is the discovery of concept graphs [35,36] in the integrated data. Concept graphs can be used to identify existing and missing concepts in a network by searching for densely connected quasi bi-partite sub-graphs. ...
The integration of heterogeneous data from various domains without the need for prefiltering prepares the ground for bisociative knowledge discoveries where attempts are made to find unexpected relations across seemingly unrelated domains. Information networks, due to their flexible data structure, lend themselves perfectly to the integration of these heterogeneous data sources. This chapter provides an overview of different types of information networks and categorizes them by identifying several key properties of information units and relations which reflect the expressiveness and thus ability of an information network to model heterogeneous data from diverse domains. The chapter progresses by describing a new type of information network known as bisociative information networks. This kind of network combines the key properties of existing networks in order to provide the foundation for bisociative knowledge discoveries. Finally based on this data structure three different patterns are described that fulfill the requirements of a bisociation by connecting concepts from seemingly unrelated domains.
... Kötter and Berthold [6] propose a new approach to extract existing concepts, or to detect missing ones, from a BisoNet by means of concept graph detection. Extracted concepts can then be used to create a higher level representation of the data, while discovered missing concepts might lead to new insights by connecting seemingly unrelated information units. ...
Heterogeneous information networks or BisoNets, as they are called in the context of bisociative knowledge discovery, are a flexible and popular form of representing data in numerous fields. Additionally, such networks can be created or derived from other types of information using, e.g., the methods given in Part II of this volume.
This part of the book describes various network algorithms for the exploration and analysis of BisoNets. Their general goal is to support and partially even automate the process of bisociation. More specific goals are to allow navigation of BisoNets by indirect and predicted relationships and by analogy, to produce explanations for discovered relationships, and to help abstract and summarise BisoNets for more effective visualisation.
... It is interesting to note that quite a few of the existing methods in the machine learning and data analysis areas can be used, frequently with only minor modifications. For instance, methods for item set mining can be applied to the detection of concept graphs [15] and measures of bisociation strength can also be derived from other approaches to model interestingness [20,22]. Bisociative Knowledge Discovery can rely to a fairly large extent on existing methods, however the way in which these methods are applied is often radically different. ...
Knowledge discovery generally focuses on finding patterns within a reasonably well connected domain of interest. In this article we outline a framework for the discovery of new connections between domains (so called bisociations), supporting the creative discovery process in a more powerful way. We motivate this approach, show the difference to classical data analysis and conclude by describing a number of different types of domain-crossing connections.
This paper considers the notion of fitness in evolutionary art and music. A taxonomy is presented of the ways in which fitness is used in such systems, with two dimensions: what the fitness function is applied to, and the basis by which the function is constructed. Papers from a large collection are classified using this taxonomy. The paper then discusses a number of ideas that have not be used for fitness evaluation in evolutionary art and which might be valuable in future developments: memory, scaffolding, connotation and web search.
Computational creativity is the application of computers to perform tasks that would be regarded as creative if performed by humans. One approach to computational creativity is to regard it as a search process, where some conceptual space is searched, and perhaps transformed, to find an outcome that would be regarded as creative. Typically, such search processes have been guided by one or more objective functions that judge how creative each solution is on one or more dimensions. This paper introduces a contrasting approach, which is search based on the idea of connotations. Rather than exploring a space constructed solely of potential outcomes, a larger space is explored consisting of such outcomes together with other relevant information. This allows us to define search processes that include a more exploratory process, out of which an outcome emerges via density of connotations. Both the general principles behind this and some specific ideas are explored.
This article proposes a new approach to extract existing (or detect missing) concepts from a loosely integrated collection of information units by means of concept graph detection. Thereby a concept graph defines a concept by a quasi bipartite sub-graph of a bigger network with the members of the concept as the first vertex partition and their shared aspects as the second vertex partition. Once the concepts have been extracted they can be used to create higher level representations of the data. Concept graphs further allow the discovery of missing concepts, which could lead to new insights by connecting seemingly unrelated information units.
