
Analysis of Images, Social Networks and Texts
Description
Alles über E-Books | Antworten auf Fragen rund um E-Books, Kopierschutz und Dateiformate finden Sie in unserem Info- & Hilfebereich.
This book constitutes the proceedings of the Fourth International Conference on Analysis of Images, Social Networks and Texts, AIST 2015, held in Yekaterinburg, Russia, in April 2015.
The 24 full and 8 short papers were carefully reviewed and selected from 140 submissions. The papers are organized in topical sections on analysis of images and videos; pattern recognition and machine learning; social network analysis; text mining and natural language processing.
More details
Other editions
Additional editions

Content
- Intro
- Preface
- Organisation
- Contents
- Invited Papers
- A Probabilistic Rating System for Team Competitions with Individual Contributions
- 1 Introduction
- 2 Problem Setting
- 3 Modeling with Unknown Individual Contributions
- 3.1 Baseline: Logistic Regression
- 3.2 Model with Latent Variables and the EM Algorithm
- 4 Experimental Evaluation
- 4.1 Dataset
- 4.2 Evaluation Metrics
- 4.3 Results
- 5 Conclusion
- References
- Sequential Hierarchical Image Recognition Based on the Pyramid Histograms of Oriented Gradients with Small Samples
- Abstract
- 1 Introduction
- 2 Classification of the PHOG with the HT-PNN
- 3 Sequential Classification of the PHOG
- 4 Experimental Results
- 5 Conclusion and Future Work
- Acknowledgements
- References
- Discerning Depression Propensity Among Participants of Suicide and Depression-Related Groups of Vk.com
- 1 Introduction
- 2 Dataset and Research Design
- 2.1 User Sampling
- 2.2 Features
- 3 Exploratory Analysis
- 3.1 Demographics
- 3.2 Social Features
- 3.3 Ego Network Features
- 4 Analysis
- 4.1 Framework
- 4.2 Model Comparison
- 4.3 Coefficient Analysis
- 4.4 Results
- 5 Conclusion
- References
- Tutorial
- Normalization of Non-standard Words with Finite State Transducers for Russian Speech Synthesis
- Abstract
- 1 Introduction
- 2 Related Work
- 3 Finite-State Transducers
- 3.1 Cardinal Numbers
- 3.2 Ordinal Numbers
- 3.3 Acronyms
- 3.4 Graphic Abbreviations
- 4 Results
- 5 Discussion
- References
- Analysis of Images and Videos
- Transform Coding Method for Hyperspectral Data: Influence of Block Characteristics to Compression Quality
- Abstract
- 1 Introduction
- 2 Generalized Transform Coding Method
- 3 Experimental Research
- 4 Conclusion
- Acknowledgements
- References
- Fréchet Filters for Color and Hyperspectral Images Filtering
- 1 Introduction
- 2 The Suboptimal Fréchet Point
- 3 Generalized Vector Aggregation
- 4 Experiments
- 5 Conclusions
- References
- Fast Global Image Denoising Algorithm on the Basis of Nonstationary Gamma-Normal Statistical Model
- 1 Introduction
- 2 Gamma-Normal Model of Hidden Random Field
- 3 The Bayesian Estimate of the Hidden Random Field
- 4 The Gauss-Seudel Procedure of Edge-Preserving Image Denoising
- 5 Experimental Results
- 6 Conclusion
- References
- Theoretical Approach to Developing Efficient Algorithms of Fingerprint Enhancement
- 1 Introduction
- 2 Proposed Approach
- 3 Coarse Field Estimation
- 4 Extraction of Cores and Deltas
- 5 Model-Based of Approximation
- 6 Curved Gabor Filtering
- 7 Conclusion
- References
- Remote Sensing Data Verification Using Model-Oriented Descriptors
- Abstract
- 1 Introduction
- 2 Problem Definition
- 3 Proposed Solution
- 4 Experiments
- 5 Conclusion
- Acknowledgements
- References
- New Bi-, Tri-, and Fourlateral Filters for Color and Hyperspectral Images Filtering
- 1 Introduction
- 2 The First Modification of Bilateral Filter
- 3 Vector Fourlateral (4-Factor) Filters
- 4 Simulation Experiments
- 5 Conclusion and Future Work
- References
- Frequency Analysis of Gradient Descent Method and Accuracy of Iterative Image Restoration
- Abstract
- 1 Introduction
- 2 Example of the Distorting Linear Operator
- 3 Frequency Analysis of Gradient Descent Method
- 4 Discrete Realization
- 5 Kernel of Linear Operator and Asymptotic Behavior of Gradient Descent Method
- 6 Linear Variation
- 7 Smoothing of Oscillating Functions
- 8 Conclusion
- Acknowledgments
- References
- Shape Matching Based on Skeletonization and Alignment of Primitive Chains
- 1 Introduction
- 2 Related Work
- 3 Comparison of Skeletons Based on Primitive Chains
- 4 Model Example of Shapes Dissimilarity Calculation
- 5 Experiments on Real-World Flavia Leaf DataSet
- 6 Conclusion
- References
- Color Image Restoration with Fuzzy Gaussian Mixture Model Driven Nonlocal Filter
- 1 Introduction
- 2 Fuzzy GMM Driven Nonlocal Means
- 2.1 Nonlocal Means
- 2.2 Homogeneity Based Patch Size Estimation
- 2.3 Fuzzy GMM for Patch Similarity via Histograms
- 3 Experimental Results
- 3.1 Setup and Parameters
- 3.2 Comparison Results
- 4 Conclusions
- References
- A Phase Unwrapping Algorithm for Interferometric Phase Images
- 1 Introduction
- 2 An Unwrapping Algorithm in 3D-space
- 3 Experimental Results
- 4 Conclusion
- References
- Robust Image Watermarking on Triangle Grid of Feature Points
- Abstract
- 1 Introduction
- 2 Feature Points Extraction
- 3 Building a Set of Primitives for Embedding
- 4 Embedding and Detection Algorithms
- 4.1 Embedding Algorithm
- 4.2 Detection Algorithm
- 5 Experimental Research
- 5.1 Robustness to Misdetection
- 5.2 Robustness to Geometrical Distortions
- 6 Conclusion
- Acknowledgements
- References
- Pattern Recognition and Machine Learning
- Traffic Flow Forecasting Algorithm Based on Combination of Adaptive Elementary Predictors
- Abstract
- 1 Introduction
- 2 Methodology
- 2.1 Notation and Basic Relations
- 2.2 Representation of the Network Graph by Feature Vector of Its Subgraphs
- 2.3 Dimension Reduction of the Subgraph Description with the Use of Spatial and Temporal Redundancy of the Traffic Data
- 2.4 Traffic Flow Prediction with the Use of Time-Series Models
- 2.5 Traffic Flow Prediction with the Use of Machine Learning Methods
- 2.6 Adaptive Linear Combination of Elementary Predictions
- 2.7 A Computational Procedure for Calculating Predictions of Traffic Flow Parameters
- 3 Test Results and Analysis
- 3.1 Accuracy Measurement
- 3.2 Computation Measurement
- 4 Conclusion and Future Work
- Acknowledgements
- References
- Analysis of the Adaptive Nature of Collaborative Filtering Techniques in Dynamic Environment
- 1 Introduction
- 2 Preliminary Concept
- 2.1 Problem Specification
- 2.2 Rating Prediction
- 2.3 Mean Absolute Error
- 2.4 Coverage
- 3 Background Study
- 4 Proposed Methodology
- 4.1 Hypothesis
- 4.2 Proposed Recommendation Life Cycle
- 5 Experimental Results
- 5.1 Experimental Configuration
- 5.2 Experimental Result
- 6 Conclusion
- References
- A Texture Fuzzy Classifier Based on the Training Set Clustering by a Self-Organizing Neural Network
- Abstract
- 1 Introduction
- 2 Pattern Representation as a Set of SOM Clusters
- 3 Fuzzy Classifier
- 4 Experiments and Results
- 5 Conclusions and Future Work
- References
- Learning Representations in Directed Networks
- 1 Introduction
- 2 Bilinear Link Model
- 3 Related Work
- 4 Experiments
- 4.1 Implementation Details
- 4.2 Graph Visualization
- 4.3 Link Prediction
- 5 Discussion and Future Work
- References
- Distorted High-Dimensional Binary Patterns Search by Scalar Neural Network Tree
- Abstract
- 1 Introduction
- 2 Problem Statement
- 3 The Point of the Algorithm
- 4 The Process of Learning
- 5 The Search Algorithm
- 6 Example of the Algorithm Operation
- 7 Estimation of the Error Probability
- 8 Estimation of the Computational Complexity
- 9 Conclusion
- 1 Appendix A
- 2 Appendix B
- References
- Hybrid Classification Approach to Decision Support for Endoscopy in Gastrointestinal Tract
- Abstract
- 1 Introduction
- 2 Analysis of Endoscopy Images
- 3 Semantic Analysis of Endoscopy Reports
- 4 Results and Discussion
- 5 Conclusions
- References
- User Similarity Computation for Collaborative Filtering Using Dynamic Implicit Trust
- 1 Introduction
- 2 Backgound and Related Work
- 3 Proposed Approach
- 3.1 Trust-Based Similarity Computation
- 3.2 Similarity Computation Module
- 3.3 Trust Computation Module
- 3.4 New Implicit Trust Measurement Method
- 3.5 Method for Tuning Constant ()
- 3.6 Combined Trust and Similarity Computation
- 4 Experimental Results
- 4.1 Dataset
- 4.2 Evaluation Metrics
- 4.3 Performance Results
- 4.4 Discussion
- 5 Conclusion
- References
- Similarity Aggregation for Collaborative Filtering
- 1 Introduction
- 2 Proposed Method
- 2.1 Computing Support Matrix
- 2.2 Finding Super Similar, Average Similar and Super Dissimilar User Pairs
- 2.3 Prediction Function
- 3 Experimental Result
- 4 Conclusion
- References
- Distributed Coordinate Descent for L1-regularized Logistic Regression
- 1 Introduction
- 2 Parallel Coordinate Descent Algorithm
- 2.1 Convergence
- 3 Scalable Software Implementation
- 4 Numerical Experiments
- 4.1 Datasets and Experimental Settings
- 4.2 Experimental Protocol for d-GLMNET
- 4.3 Experimental Protocol for Distributed Online Learning via Truncated Gradient
- 4.4 Results
- References
- Social Network Analysis
- Building Profiles of Blog Users Based on Comment Graph Analysis: The Habrahabr.ru Case
- 1 Introduction
- 2 Related Work
- 3 Definition of a Comment Graph
- 3.1 Community Graph
- 3.2 Blog Post Graph
- 4 User Clustering Algorithm
- 4.1 User Attributes
- 4.2 Clustering
- 5 Habrahabr Data Set
- 6 Results
- 7 Visualisation Tool
- 8 Conclusion
- References
- Formation and Evolution Mechanisms in Online Network of Students: The Vkontakte Case
- 1 Introduction and Related Works
- 2 Network Growth Models
- 3 Data
- 3.1 Description of the Case
- 3.2 Data Collection Procedure
- 4 Results
- 4.1 Descriptive Statistics
- 4.2 Modeling Results
- 4.3 Conclusion
- References
- Large-Scale Parallel Matching of Social Network Profiles
- 1 Introduction
- 2 Related Work
- 2.1 Profile Matching
- 2.2 Name Similarity Matching
- 3 Dataset
- 3.1 VKontakte
- 3.2 Facebook
- 3.3 Test Data
- 3.4 Name Romanisation
- 4 Profile Matching Algorithm
- 4.1 Candidate Generation
- 4.2 Candidate Ranking
- 4.3 Best Candidate Selection
- 5 Results and Discussion
- 6 Conclusion
- References
- Identification of Autopoietic Communication Patterns in Social and Economic Networks
- 1 Introduction
- 2 Economic Agent Communications
- 3 Balanced Matrix Development
- 4 Social Agent Communications
- 5 Conclusions
- References
- Text Mining and Natural Language Processing
- A Heuristic Strategy for Extracting Terms from Scientific Texts
- Abstract
- 1 Introduction
- 2 Scientific Terms and Their Lexico-Syntactic Patterns
- 3 Testing Term Extraction Procedures
- 4 Strategy for Combining Sets of Extracted Terms
- 5 Conclusion
- Acknowledgements
- References
- Text Analysis with Enhanced Annotated Suffix Trees: Algorithms and Implementation
- 1 Introduction
- 2 Annotated Suffix Trees
- 3 Enhanced AST Method with Suffix Arrays
- 4 EAST Software Package
- 5 Experimental Results
- 6 Conclusion
- References
- Morphological Analyzer and Generator for Russian and Ukrainian Languages
- 1 Introduction
- 2 Software Architecture
- 3 Analysis of Vocabulary Words
- 3.1 Morphological Analysis and Generation
- 3.2 Working with "?" and "?" Characters Efficiently
- 4 Analysis of Out-of-Vocabulary Words
- 4.1 Common Prefixes Removal
- 4.2 Words Ending with Other Dictionary Words
- 4.3 Endings Matching
- 4.4 Words with a Hyphen
- 4.5 Other Tokens
- 4.6 Morphological Generation of Out of Vocabulary Words
- 5 Probability Estimation
- 6 Evaluation
- 7 Conclusion and Future Plans
- References
- Semantic Role Labeling for Russian Language Based on Russian FrameBank
- 1 Introduction
- 2 FrameBank-Based SRL for Russian
- 2.1 Source Data
- 2.2 Preprocessing
- 2.3 SRL Engine
- 3 Evaluation and Results
- 4 Related Work
- 5 Conclusion
- References
- Supervised Approach to Finding Most Frequent Senses in Russian
- 1 Introduction
- 2 Related Work
- 3 RuThes Linguistic Ontology
- 4 Manual Analysis of Sense Distribution
- 5 Supervised Estimation of Most Frequent Sense
- 6 Conclusion
- References
- FrameBank: A Database of Russian Lexical Constructions
- Abstract
- 1 Background
- 2 Dictionary of Construction Patterns
- 3 Corpus Annotation
- 4 Semantic Roles
- 5 Non-core Elements
- 6 Construction Grapher
- 6.1 Formal Changes
- 6.2 Semantic Changes
- 7 Future Prospects
- References
- TagBag: Annotating a Foreign Language Lexical Resource with Pictures
- 1 Introduction
- 2 Related Work
- 3 TagBag
- 4 Evaluation
- 4.1 Experimental Setup
- 4.2 Deriving a Bilingual Dictionary
- 4.3 Questionnaire
- 4.4 Results
- 5 Discussion
- 5.1 Controversial Mappings
- 5.2 Threshold
- 5.3 Similarity Measures
- 6 Conclusion
- References
- BigARTM: Open Source Library for Regularized Multimodal Topic Modeling of Large Collections
- 1 Introduction
- 2 Multimodal Regularized Topic Model
- 3 Online Topic Modeling
- 4 BigARTM Architecture
- 5 Experiments
- 6 Conclusions
- References
- Industry Talk
- ATM Service Cost Optimization Using Predictive Encashment Strategy
- 1 Introduction
- 2 Related Work
- 3 Business Problem Identification
- 4 Encashment Strategy Description
- 4.1 Technical Requirements
- 4.2 Device Thresholds
- 4.3 Encashment Amount Determination
- 5 Solution Architecture
- 5.1 Data Processing and Feature Extraction
- 5.2 Predictive Model Evaluation
- 5.3 Encashment Plan Provision
- 6 Results
- 7 Discussion
- References
- Industry Papers
- Comparison of Deep Learning Libraries on the Problem of Handwritten Digit Classification
- 1 Introduction
- 2 The Problem of Deep Learning Libraries Comparison
- 3 Deep Learning Libraries, Capabilities and Features
- 3.1 Deep Learning Frameworks
- 3.2 Caffe Library
- 3.3 Pylearn2 Library
- 3.4 Torch Library
- 3.5 Theano Library
- 4 Comparison of Libraries by the Example of Handwritten Digit Classification
- 4.1 The Handwritten Digit Classification Problem
- 4.2 Infrastructure
- 4.3 Learning Parameters
- 4.4 Dataset
- 4.5 Performance and Accuracy Analysis
- 4.6 Comparison of Chosen Libraries
- 5 Conclusion
- References
- Methods of Localization of Some Anthropometric Features of Face
- Abstract
- 1 Introduction
- 2 Solving the Task of Localization of the Potential Regions
- 3 Training of a Classifier for Face Feature Localization
- 4 Finding the Resultant Square Out of the Array of Potential Regions
- 5 Modification of Localization Algorithm by Adding More Classifiers
- 6 Results of the Experiment
- 7 Conclusion
- References
- Ontological Representation of Networks for IDS in Cyber-Physical Systems
- 1 Introduction
- 2 Intrusion Detection Systems
- 3 Architecture
- 4 Evaluation
- 5 Related Works
- 6 Conclusion
- References
- Determination of the Relative Position of Space Vehicles by Detection and Tracking of Natural Visual Features with the Existing TV-Cameras
- Abstract
- 1 Introduction
- 2 Statement of the Problem
- 3 Algorithm Description
- 3.1 Camera Model
- 3.2 Target Model
- 3.3 Analog Signal Distortion and Camera Model Correction
- 3.4 Global and Local Image Normalization
- 3.5 Parallel Detection and Tracking
- 3.6 The Relative Pose Estimation
- 3.7 Self-Evaluation of the Solutions Based on Reprojection Error
- 3.8 Distance-Based Sets of Points Switching
- 4 System Implementation
- 5 Evaluation
- 5.1 Modules Evaluation
- 5.2 Evaluation Using Recordings of Rendezvous and the ``Course'' Telemetry
- 5.3 Real-Time Experiments at the Mission Control Center
- 6 Conclusion
- References
- Implementation of Agile Concepts in Recommender Systems for Data Processing and Analyses
- Abstract
- 1 Introduction
- 2 Agile Concepts for DPAS and RS
- 3 Common and Agile Features of DPA RS
- 4 DPA RS Information Model
- 5 DPA RS Architecture
- 6 DPA RS Implementation and Usage
- 7 Conclusion
- References
- Author Index
System requirements
File format: PDF
Copy protection: Watermark-DRM (Digital Rights Management)
System requirements:
- Computer (Windows; MacOS X; Linux): Use the free software Adobe Reader, Adobe Digital Editions, or any other PDF viewer of your choice (see eBook Help).
- Tablet/Smartphone (Android; iOS): Install the free app Adobe Digital Editions or another reading app for eBooks, e.g., PocketBook (see eBook Help).
- E-reader: Bookeen, Kobo, Pocketbook, Sony, Tolino and many more (only limited: Kindle).
The file format PDF always displays a book page identically on any hardware. This makes PDF suitable for complex layouts such as those used in textbooks and reference books (images, tables, columns, footnotes). Unfortunately, on the small screens of e-readers or smartphones, PDFs are rather annoying, requiring too much scrolling.
This eBook uses Watermark-DRM, a „soft” copy protection. This means that there are no technical restrictions to prevent illegal distribution. However, there is a personalised watermark embedded in the eBook that can be used to identify the purchaser of the eBook in the event of misuse and to provide evidence for legal purposes.
For more information, see our eBook Help page.