DOCSLIB.ORG

Adjacency List Oriented Relational Fact Extraction Via Adaptive Multi

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Adjacency List Oriented Relational Fact Extraction Via Adaptive Multi Adjacency List Oriented Relational Fact Extraction via Adaptive Multi-task Learning Fubang Zhao1∗, Zhuoren Jiang2∗y, Yangyang Kang1, Changlong Sun1, Xiaozhong Liu3 1Alibaba Group, Hangzhou, China 2School of Public Affairs, Zhejiang University, Hangzhou, China 3School of Informatics, Computing and Engineering, IUB, Bloomington, USA [email protected], [email protected] [email protected], [email protected] [email protected] Abstract Relational fact extraction aims to extract se- mantic triplets from unstructured text. In this work, we show that all of the relational fact extraction models can be organized accord- ing to a graph-oriented analytical perspective. An efficient model, aDjacency lIst oRiented rElational faCT(DIRECT), is proposed based on this analytical framework. To alleviate challenges of error propagation and sub-task loss equilibrium, DIRECT employs a novel adaptive multi-task learning strategy with dy- namic sub-task loss balancing. Extensive ex- periments are conducted on two benchmark Figure 1: Example of exploring the relational fact ex- datasets, and results prove that the proposed traction task from the perspective of directed graph rep- model outperforms a series of state-of-the-art resentation method as output data structure. (SoTA) models for relational triplet extraction. 1 Introduction et al., 2019), and PNDec (Nayak and Ng, 2020), Relational fact extraction, as an essential NLP fall into this category. task, is playing an increasingly important role in Edge list is a simple and space-efficient way to knowledge graph construction (Han et al., 2019; represent a graph (Arifuzzaman and Khan, 2015). Distiawan et al., 2019). It aims to extract rela- However, there are three problems. First, the tional triplet from the text. A relational triplet triplet overlapping problem (Zeng et al., 2018). is in the form of (subject; relation; object) or For instance, as shown in Figure1, for triplets (s; r; o) (Zeng et al., 2019). While various prior (Obama, nationality, USA) and (Obama, presi- models proposed for relational fact extraction, few dent of, USA), there are two types of relations be- of them analyze this task from the perspective of tween the “Obama” and “USA”. If the model only output data structure. generates one sequence from the text (Zheng et al., arXiv:2106.01559v1 [cs.CL] 3 Jun 2021 As shown in Figure1, the relational fact extrac- 2017), it may fail to identify the multi-relation be- tion can be characterized as a directed graph con- tween entities. Second, to overcome the triplet over- struction task, where graph representation flexibil- lapping problem, the model may have to extract ity and heterogeneity accompany additional bene- the triplet element repeatedly (Zeng et al., 2018), faction. In practice, there are three common ways which will increase the extraction cost. Third, there to represent graphs (Gross and Yellen, 2005): could be an ordering problem (Zeng et al., 2019): Edge List is utilized to predict a sequence of for multiple triplets, the extraction order could in- triplets (edges). The recent sequence-to-sequence fluence the model performance. based models, such as NovelTagging (Zheng et al., Adjacency Matrices are used to predict ma- 2017), CopyRE (Zeng et al., 2018), CopyRL (Zeng trices that represent exactly which entities (ver- ∗These two authors contributed equally to this research. tices) have semantic relations (edges) between y Zhuoren Jiang is the corresponding author them. Most early works, which take a pipeline ap- proach (Zelenko et al., 2003; Zhou et al., 2005), be- lem (Chen et al., 2018; Sener and Koltun, 2018) long to this category. These models first recognize could compromise its performance. all entities in text and then perform relation classi- Based on the analysis from the perspective of fication for each entity pair. The subsequent neural output data structure, we propose a novel solution, network-based models (Bekoulis et al., 2018; Dai aDjacency lIst oRiented rElational faCT extraction et al., 2019), that attempt to extract entities and model (DIRECT), with the following advantages: relations jointly, can also be classified into this cat- • For efficiency, DIRECT is a fully adjacency list egory. oriented model, which consists of a shared BERT Compared to edge list, adjacency matrices have encoder, the Pointer-Network based subject and ob- better relation (edge) searching efficiency (Arifuz- ject extractors, and a relation classification module. zaman and Khan, 2015). Furthermore, adjacency In Section 3.4, we provide a detailed comparative matrices oriented models is able to cover differ- analysis2 to demonstrate the efficiency of the pro- ent overlapping cases (Zeng et al., 2018) for rela- posed method. tional fact extraction task. But the space cost of • From the performance viewpoint, to address this approach can be expensive. For most cases, the sub-task error propagation and sub-task loss balanc- output matrices are very sparse. For instance, for ing problems, DIRECT employs a novel adaptive a sentence with n tokens, if there are m kinds of multi-task learning strategy with the dynamic sub- relations, the output space is n · n · m, which can task loss balancing approach. In Section 3.2 and be costly for graph representation efficiency. This 3.3, the empirical experimental results demonstrate phenomenon is also illustrated in Figure1. DIRECT can achieve the state-of-the-art perfor- Adjacency List is designed to predict an array mance of relational fact extraction task, and the of linked lists that serves as a representation of a adaptive multi-task learning strategy did play a pos- graph. As depicted in Figure1, in the adjacency itive role in improving the task performance. list, each vertex v (key) points to a list (value) con- The major contributions of this paper can be taining all other vertices connected to v by sev- summarized as follows: eral edges. Adjacency list is a hybrid graph rep- 1. We refurbish the relational fact extraction resentation between edge list and adjacency ma- problem by leveraging an analytical framework trices (Gross and Yellen, 2005), which can bal- of graph-oriented output structure. To the best of ance space and searching efficiency1. Due to the our knowledge, this is a pioneer investigation to structural characteristic of the adjacency list, this explore the output data structure of relational fact type of model usually adopts a cascade fashion to extractions. identify subject, object, and relation sequentially. 2. We propose a novel solution, DIRECT3, For instance, the recent state-of-the-art model Cas- which is a fully adjacency list oriented model with Rel (Wei et al., 2020) can be considered as an ex- a novel adaptive multi-task learning strategy. emplar. It utilizes a two-step framework to rec- 3. Through extensive experiments on two bench- ognize the possible object(s) of a given subject mark datasets3, we demonstrate the efficiency and under a specific relation. However, CasRel is not efficacy of DIRECT. The proposed DIRECT out- fully adjacency list oriented: in the first step, it performs the state-of-the-art baseline models. use subject as the key; while in the second step, it predicts (relation; object) pairs using adjacency 2 The DIRECT Framework matrix representation. In this section, we will introduce the framework Despite its considerable potential, the cascade of the proposed DIRECT model, which includes fashion of adjacency list oriented model may cause a shared BERT encoder and three output layers: problems of sub-task error propagation (Shen et al., subject extraction, object extraction, and relation 2019), i.e., errors from ancestor sub-tasks may ac- classification. As shown in Figure2, DIRECT is cumulate to threaten downstream ones, and sub- fully adjacency list oriented. The input sentence tasks can hardly share supervision signals. Multi- is firstly fed into the subject extraction module to task learning (Caruana, 1997) can alleviate this 2Theoretical representation efficiency analysis of graph problem, however, the sub-task loss balancing prob- representative models are described in Appendix section 6.4. 3To help other scholars reproduce the experiment out- 1More detailed complexity analyses of different graph rep- come, we will release the code and datasets via GitHub: resentations are provided in Appendix section 6.3. https://github.com/fyubang/direct-ie. Figure 2: An overview of the proposed DIRECT framework extract all subjects. Then each extracted subject 2.2 Subject and Object Extraction is concatenated with the sentence, and fed into The subject and object extraction modules are moti- the object extraction module to extract all objects, vated by the Pointer-Network (Vinyals et al., 2015) which can form a set of subject-object pairs. Fi- architecture, which are widely used in Machine nally, the subject-object pair is concatenated with Reading Comprehension (MRC) (Rajpurkar et al., sentence, and fed into the relation classification 2016) task. Different from MRC task that only module to get the relations between them. For needs to extract a single span, the subject and object balancing the weights of sub-task losses and to im- extractions need to extract multiple spans. There- prove the global task performance, three modules fore, in the training phase, we replace softmax share the BERT encoder layer and are trained with function with sigmoid function for the activation an adaptive multi-task learning strategy. function of the output layer, and replace cross en- tropy (CE) (Goodfellow et al., 2016) with binary 2.1 Shared BERT Encoder cross entropy (BCE) (Luc et al., 2016) for the loss In the DIRECT framework, the encoder is used to function. Specifically, we will perform indepen- extract the semantic features from the inputs for dent binary classifications for each token twice to three modules. As aforementioned, we employ the indicate whether the current token is the start or the BERT (Devlin et al., 2019) as the shared encoder to end of a span. The probability of a token to be start make use of its pre-trained knowledge and attention or end is as follows: mechanism.
Load more

Recommended publications
  • Practical Parallel Hypergraph Algorithms
    Practical Parallel Hypergraph Algorithms Julian Shun [email protected] MIT CSAIL Abstract v While there has been signicant work on parallel graph pro- 0 cessing, there has been very surprisingly little work on high- e0 performance hypergraph processing. This paper presents v0 v1 v1 a collection of ecient parallel algorithms for hypergraph processing, including algorithms for betweenness central- e1 ity, maximal independent set, k-core decomposition, hyper- v2 trees, hyperpaths, connected components, PageRank, and v2 v3 e single-source shortest paths. For these problems, we either 2 provide new parallel algorithms or more ecient implemen- v3 tations than prior work. Furthermore, our algorithms are theoretically-ecient in terms of work and depth. To imple- (a) Hypergraph (b) Bipartite representation ment our algorithms, we extend the Ligra graph processing Figure 1. An example hypergraph representing the groups framework to support hypergraphs, and our implementa- , , , , , , and , , and its bipartite repre- { 0 1 2} { 1 2 3} { 0 3} tions benet from graph optimizations including switching sentation. between sparse and dense traversals based on the frontier size, edge-aware parallelization, using buckets to prioritize processing of vertices, and compression. Our experiments represented as hyperedges, can contain an arbitrary number on a 72-core machine and show that our algorithms obtain of vertices. Hyperedges correspond to group relationships excellent parallel speedups, and are signicantly faster than among vertices (e.g., a community in a social network). An algorithms in existing hypergraph processing frameworks. example of a hypergraph is shown in Figure 1a. CCS Concepts • Computing methodologies → Paral- Hypergraphs have been shown to enable richer analy- lel algorithms; Shared memory algorithms.
    [Show full text]
  • Practical Parallel Hypergraph Algorithms
    Practical Parallel Hypergraph Algorithms Julian Shun [email protected] MIT CSAIL Abstract v0 While there has been significant work on parallel graph pro- e0 cessing, there has been very surprisingly little work on high- v0 v1 v1 performance hypergraph processing. This paper presents a e collection of efficient parallel algorithms for hypergraph pro- 1 v2 cessing, including algorithms for computing hypertrees, hy- v v 2 3 e perpaths, betweenness centrality, maximal independent sets, 2 v k-core decomposition, connected components, PageRank, 3 and single-source shortest paths. For these problems, we ei- (a) Hypergraph (b) Bipartite representation ther provide new parallel algorithms or more efficient imple- mentations than prior work. Furthermore, our algorithms are Figure 1. An example hypergraph representing the groups theoretically-efficient in terms of work and depth. To imple- fv0;v1;v2g, fv1;v2;v3g, and fv0;v3g, and its bipartite repre- ment our algorithms, we extend the Ligra graph processing sentation. framework to support hypergraphs, and our implementations benefit from graph optimizations including switching between improved compared to using a graph representation. Unfor- sparse and dense traversals based on the frontier size, edge- tunately, there is been little research on parallel hypergraph aware parallelization, using buckets to prioritize processing processing. of vertices, and compression. Our experiments on a 72-core The main contribution of this paper is a suite of efficient machine and show that our algorithms obtain excellent paral- parallel hypergraph algorithms, including algorithms for hy- lel speedups, and are significantly faster than algorithms in pertrees, hyperpaths, betweenness centrality, maximal inde- existing hypergraph processing frameworks.
    [Show full text]
  • Adjacency Matrix
    CSE 373 Graphs 3: Implementation reading: Weiss Ch. 9 slides created by Marty Stepp http://www.cs.washington.edu/373/ © University of Washington, all rights reserved. 1 Implementing a graph • If we wanted to program an actual data structure to represent a graph, what information would we need to store? for each vertex? for each edge? 1 2 3 • What kinds of questions would we want to be able to answer quickly: 4 5 6 about a vertex? about edges / neighbors? 7 about paths? about what edges exist in the graph? • We'll explore three common graph implementation strategies: edge list , adjacency list , adjacency matrix 2 Edge list • edge list : An unordered list of all edges in the graph. an array, array list, or linked list • advantages : 1 2 3 easy to loop/iterate over all edges 4 5 6 • disadvantages : hard to quickly tell if an edge 7 exists from vertex A to B hard to quickly find the degree of a vertex (how many edges touch it) 0 1 2 3 4 56 7 8 (1, 2) (1, 4) (1, 7) (2, 3) 2, 5) (3, 6)(4, 7) (5, 6) (6, 7) 3 Graph operations • Using an edge list, how would you find: all neighbors of a given vertex? the degree of a given vertex? whether there is an edge from A to B? 1 2 3 whether there are any loops (self-edges)? • What is the Big-Oh of each operation? 4 5 6 7 0 1 2 3 4 56 7 8 (1, 2) (1, 4) (1, 7) (2, 3) 2, 5) (3, 6)(4, 7) (5, 6) (6, 7) 4 Adjacency matrix • adjacency matrix : An N × N matrix where: the non-diagonal entry a[i,j] is the number of edges joining vertex i and vertex j (or the weight of the edge joining vertex i and vertex j).
    [Show full text]
  • Introduction to Graphs
    Multidimensional Arrays & Graphs CMSC 420: Lecture 3 Mini-Review • Abstract Data Types: • Implementations: • List • Linked Lists • Stack • Circularly linked lists • Queue • Doubly linked lists • Deque • XOR Doubly linked lists • Dictionary • Ring buffers • Set • Double stacks • Bit vectors Techniques: Sentinels, Zig-zag scan, link inversion, bit twiddling, self- organizing lists, constant-time initialization Constant-Time Initialization • Design problem: - Suppose you have a long array, most values are 0. - Want constant time access and update - Have as much space as you need. • Create a big array: - a = new int[LARGE_N]; - Too slow: for(i=0; i < LARGE_N; i++) a[i] = 0 • Want to somehow implicitly initialize all values to 0 in constant time... Constant-Time Initialization means unchanged 1 2 6 12 13 Data[] = • Access(i): if (0≤ When[i] < count and Where[When[i]] == i) return Where[] = 6 13 12 Count = 3 Count holds # of elements changed Where holds indices of the changed elements. When[] = 1 3 2 When maps from index i to the time step when item i was first changed. Access(i): if 0 ≤ When[i] < Count and Where[When[i]] == i: return Data[i] else: return DEFAULT Multidimensional Arrays • Often it’s more natural to index data items by keys that have several dimensions. E.g.: • (longitude, latitude) • (row, column) of a matrix • (x,y,z) point in 3d space • Aside: why is a plane “2-dimensional”? Row-major vs. Column-major order • 2-dimensional arrays can be mapped to linear memory in two ways: 1 2 3 4 5 1 2 3 4 5 1 1 2 3 4 5 1 1 5 9 13 17 2 6 7 8 9 10 2 2 6 10 14 18 3 11 12 13 14 15 3 3 7 11 15 19 4 16 17 18 19 20 4 4 8 12 16 20 Row-major order Column-major order Addr(i,j) = Base + 5(i-1) + (j-1) Addr(i,j) = Base + (i-1) + 4(j-1) Row-major vs.
    [Show full text]
  • 9 the Graph Data Model
    CHAPTER 9 ✦ ✦ ✦ ✦ The Graph Data Model A graph is, in a sense, nothing more than a binary relation. However, it has a powerful visualization as a set of points (called nodes) connected by lines (called edges) or by arrows (called arcs). In this regard, the graph is a generalization of the tree data model that we studied in Chapter 5. Like trees, graphs come in several forms: directed/undirected, and labeled/unlabeled. Also like trees, graphs are useful in a wide spectrum of problems such as com- puting distances, finding circularities in relationships, and determining connectiv- ities. We have already seen graphs used to represent the structure of programs in Chapter 2. Graphs were used in Chapter 7 to represent binary relations and to illustrate certain properties of relations, like commutativity. We shall see graphs used to represent automata in Chapter 10 and to represent electronic circuits in Chapter 13. Several other important applications of graphs are discussed in this chapter. ✦ ✦ ✦ ✦ 9.1 What This Chapter Is About The main topics of this chapter are ✦ The definitions concerning directed and undirected graphs (Sections 9.2 and 9.10). ✦ The two principal data structures for representing graphs: adjacency lists and adjacency matrices (Section 9.3). ✦ An algorithm and data structure for finding the connected components of an undirected graph (Section 9.4). ✦ A technique for finding minimal spanning trees (Section 9.5). ✦ A useful technique for exploring graphs, called “depth-first search” (Section 9.6). 451 452 THE GRAPH DATA MODEL ✦ Applications of depth-first search to test whether a directed graph has a cycle, to find a topological order for acyclic graphs, and to determine whether there is a path from one node to another (Section 9.7).
    [Show full text]
  • List Representation: Adjacency List – N Rows of the Adjacency Matrix Are
    List Representation: Adjacency List { n rows of the adjacency matrix are repre- sented as n linked lists { Declare an Array of size n e.g. A[1:::n] of Lists { A[i] is a pointer to the edges in E(G) starting at vertex i. Undirected graph: data in each list cell is a number indicating the adjacent vertex Weighted graph: data in each list cell is a pair (vertex, weight) 1 a b c e a [b,2] [c,6] [e,4] b a b [a,2] c a e d c [a,6] [e,3] [d,1] d c d [c,1] e a c e [a,4] [c,3] Properties: { the degree of any node is the number of elements in the list { O(e) to check for adjacency for e edges. { O(n + e) space for graph with n vertices and e edges. 2 Search and Traversal Techniques Trees and Graphs are models on which algo- rithmic solutions for many problems are con- structed. Traversal: All nodes of a tree/graph are exam- ined/evaluated, e.g. | Evaluate an expression | Locate all the neighbours of a vertex V in a graph Search: only a subset of vertices(nodes) are examined, e.g. | Find the first token of a certain value in an Abstract Syntax Tree. 3 Types of Traversals Binary Tree traversals: | Inorder, Postorder, Preorder General Tree traversals: | With many children at each node Graph Traversals: | With Trees as a special case of a graph 4 Binary Tree Traversal Recall a Binary tree | is a tree structure in which there can be at most two children for each parent | A single node is the root | The nodes without children are leaves | A parent can have a left child (subtree) and a right child (subtree) A node may be a record of information | A node is visited when it is considered in the traversal | Visiting a node may involve computation with one or more of the data fields at the node.
    [Show full text]
  • Package 'Igraph'
    Package ‘igraph’ February 28, 2013 Version 0.6.5-1 Date 2013-02-27 Title Network analysis and visualization Author See AUTHORS file. Maintainer Gabor Csardi <[email protected]> Description Routines for simple graphs and network analysis. igraph can handle large graphs very well and provides functions for generating random and regular graphs, graph visualization,centrality indices and much more. Depends stats Imports Matrix Suggests igraphdata, stats4, rgl, tcltk, graph, Matrix, ape, XML,jpeg, png License GPL (>= 2) URL http://igraph.sourceforge.net SystemRequirements gmp, libxml2 NeedsCompilation yes Repository CRAN Date/Publication 2013-02-28 07:57:40 R topics documented: igraph-package . .5 aging.prefatt.game . .8 alpha.centrality . 10 arpack . 11 articulation.points . 15 as.directed . 16 1 2 R topics documented: as.igraph . 18 assortativity . 19 attributes . 21 autocurve.edges . 23 barabasi.game . 24 betweenness . 26 biconnected.components . 28 bipartite.mapping . 29 bipartite.projection . 31 bonpow . 32 canonical.permutation . 34 centralization . 36 cliques . 39 closeness . 40 clusters . 42 cocitation . 43 cohesive.blocks . 44 Combining attributes . 48 communities . 51 community.to.membership . 55 compare.communities . 56 components . 57 constraint . 58 contract.vertices . 59 conversion . 60 conversion between igraph and graphNEL graphs . 62 convex.hull . 63 decompose.graph . 64 degree . 65 degree.sequence.game . 66 dendPlot . 67 dendPlot.communities . 68 dendPlot.igraphHRG . 70 diameter . 72 dominator.tree . 73 Drawing graphs . 74 dyad.census . 80 eccentricity . 81 edge.betweenness.community . 82 edge.connectivity . 84 erdos.renyi.game . 86 evcent . 87 fastgreedy.community . 89 forest.fire.game . 90 get.adjlist . 92 get.edge.ids . 93 get.incidence . 94 get.stochastic .
    [Show full text]
  • An XML-Based Description of Structured Networks
    Massimo Ancona, Walter Cazzola, Sara Drago, and Francesco Guido. An XML-Based Description of Structured Networks. In Proceedings of Interna- tional Conference Communications 2004, pages 401–406, Bucharest, Romania, June 2004. IEEE Press. An XML-Based Description of Structured Networks Massimo Ancona£ Walter Cazzola† Sara Drago£ Francesco Guido‡ £ DISI University of Genova, e-mail: fancona,[email protected] † DICo University of Milano, e-mail: [email protected] ‡ Marconi Selenia Communication, e-mail: [email protected] Abstract In this paper we present an XML-based formalism to describe hierarchically organized communication networks. After a short overview of existing graph description languages, we discuss the advantages and disadvantages of their application in network optimization. We conclude by extending one of these formalisms with features supporting the description of the relationship between the optimized logical layout of a network and its physical counterpart. Elements for describing traffic parameters are also given. 1 Introduction The problem of network design and optimization has a strategical importance especially for military networks. In this case, design for fault tolerance and security plays a role more and more crucial than the equivalent importance required for commercial and research networks. The reengineering of a large communication network is a complex problem, which consists of different aspects that can be strongly affected by the way of describing data. The plainest way to describe a communication network is to model the relationship among sites and links by means of a weighted undirected graph, but unless some more assumptions are taken, this approach can raise several problems. A first issue is encountered when an analysis and a visualization of the network has to be realized: practical networks include hundreds or often thousands of nodes and links, so that even a simple description and documentation of the network structure is hard to maintain and update.
    [Show full text]
  • Note: Nonlinear Data Structures Do Not Have Their Elements in a Sequence
    CS231 - Fall 2017 Graphs • Graph is a non-linear data structure (like a tree). Note: nonlinear data structures do not have their elements in a sequence. • But trees are somewhat orderly. Root is connected to its children. The children are connected to their children…all the way to the leaves. • Graph can be linked in any pattern. • A graph consists of vertices (nodes) and edges (arcs). o Name the vertices: {A,B,C,D,E} o Name the edges: {(A,B), (A,E), (A,D),…} • Applications: o roads between cities o communication networks o oil pipelines o personal relationships between people o interval graphs o rooms in caves with passages between them • Directed vs. Undirected o Graphs can be directed or undirected. o In undirected graphs, edges have no specific direction. Edges are always “two-way” o In directed graphs (also called digraphs), edges have a direction 1 o A self-edge a.k.a. a loop is an edge of the form (a,a) • Graph Connectivity o A graph is connected when there is a path between every pair of vertices. o In a connected graph, there are no unreachable vertices. o A graph that is not connected is disconnected. o A graph G is said to be disconnected if there exist two nodes in G such that no path in G has those nodes as endpoints. • Weighted graphs o In a weighed graph, each edge has a weight a.k.a. cost o Some graphs allow negative weights; some not • Paths and Cycles 2 o Path from one vertex to another is a sequence of vertices, each a neighbor of the previous one.
    [Show full text]
  • Chapter 12 Graphs and Their Representation
    Chapter 12 Graphs and their Representation 12.1 Graphs and Relations Graphs (sometimes referred to as networks) offer a way of expressing relationships between pairs of items, and are one of the most important abstractions in computer science. Question 12.1. What makes graphs so special? What makes graphs special is that they represent relationships. As you will (or might have) discover (discovered already) relationships between things from the most abstract to the most concrete, e.g., mathematical objects, things, events, people are what makes everything inter- esting. Considered in isolation, hardly anything is interesting. For example, considered in isolation, there would be nothing interesting about a person. It is only when you start consid- ering his or her relationships to the world around, the person becomes interesting. Challenge yourself to try to find something interesting about a person in isolation. You will have difficulty. Even at a biological level, what is interesting are the relationships between cells, molecules, and the biological mechanisms. Question 12.2. Trees captures relationships too, so why are graphs more interesting? Graphs are more interesting than other abstractions such as tree, which can also represent certain relationships, because graphs are more expressive. For example, in a tree, there cannot be cycles, and multiple paths between two nodes. Question 12.3. What do we mean by a “relationship”? Can you think of a mathematical way to represent relationships. 231 232 CHAPTER 12. GRAPHS AND THEIR REPRESENTATION Alice Bob Josefa Michel Figure 12.1: Friendship relation f(Alice, Bob), (Bob, Alice), (Bob, Michel), (Michel, Bob), (Josefa, Michel), (Michel, Josefa)g as a graph.
    [Show full text]
  • Data Structure for Representing a Graph: Combination of Linked List and Hash Table
    Data structure for representing a graph: combination of linked list and hash table Maxim A. Kolosovskiy Altai State Technical University, Russia [email protected] Abstract In this article we discuss a data structure, which combines advantages of two different ways for representing graphs: adjacency matrix and collection of adjacency lists. This data structure can fast add and search edges (advantages of adjacency matrix), use linear amount of memory, let to obtain adjacency list for certain vertex (advantages of collection of adjacency lists). Basic knowledge of linked lists and hash tables is required to understand this article. The article contains examples of implementation on Java. 1. Introduction There are two most common ways for representing a graph: Adjacency matrix Collection of adjacency lists Let’s look at comparison table of these ways: Adjacency Collection of matrix adjacency lists Memory complexity (optimal – O(|E|)) O(|V|2) O(|E|) Add new edge (optimal – O(1)) O(1) O(1) Remove edge (optimal – O(1)) O(1) O(|K|) Search edge (optimal – O(1)) O(1) O(|K|) Enumeration of the vertices adjacent to u O(|V|) O(|K|) (optimal – O(|K|)) Table 1. Comparison table of two ways for representing graphs. V – set of vertices, E – set of edges, K – set of vertices adjacent to vertex u. We assume that |V|<|E|. Bold font – optimal complexity. Operations with a graph represented by an adjacency matrix are faster. But if a graph is large we can’t use such big matrix to represent a graph, so we should use collection of adjacency lists, which is more compact.
    [Show full text]
  • Analyzing the Facebook Friendship Graph
    Analyzing the Facebook friendship graph S. Catanese1, P. De Meo2, E. Ferrara3, G. Fiumara1 1Dept. of Physics, Informatics Section, University of Messina 2Dept. of Computer Sciences, Vrije Universiteit Amsterdam 3Dept. of Mathematics, University of Messina 1st International Workshop on Mining the Future Internet 20 September 2010, Berlin Catanese, De Meo, Ferrara, Fiumara () Analyzing the Facebook friendship graph MIFI 2010, 20/09/2010, Berlin 1 / 42 Outline 1 Motivation Main objective The Basic Problem Classic Work 2 Our Results/Contribution Data Extraction and Cleaning Data Analysis Main Results 3 Future Issues Catanese, De Meo, Ferrara, Fiumara () Analyzing the Facebook friendship graph MIFI 2010, 20/09/2010, Berlin 2 / 42 Outline 1 Motivation Main objective The Basic Problem Classic Work 2 Our Results/Contribution Data Extraction and Cleaning Data Analysis Main Results 3 Future Issues Catanese, De Meo, Ferrara, Fiumara () Analyzing the Facebook friendship graph MIFI 2010, 20/09/2010, Berlin 3 / 42 Main objective Analyze the Facebook friendship graph using: I a wrapper (for extraction, cleaning and normalization of data) I a tool for graph visualization and analysis I developed by some of us Catanese, De Meo, Ferrara, Fiumara () Analyzing the Facebook friendship graph MIFI 2010, 20/09/2010, Berlin 4 / 42 Main objective Analyze the Facebook friendship graph using: I a wrapper (for extraction, cleaning and normalization of data) I a tool for graph visualization and analysis I developed by some of us Catanese, De Meo, Ferrara, Fiumara
    [Show full text]