Artificial Intelligence and Natural Language

Name: Artificial Intelligence and Natural Language | 6th Conference, AINL 2017, St. Petersburg, Russia, September 20-23, 2017, Revised Selected Papers
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

6th Conference, AINL 2017, St. Petersburg, Russia, September 20-23, 2017, Revised Selected Papers

Andrey Filchenkov Lidia Pivovarova Jan Zizka(Herausgeber*in)

Springer (Verlag)

Erschienen am 28. November 2017

XI, 305 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-319-71746-3 (ISBN)

53,49 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Preface
Organization
Contents
Social Interaction Analysis
Semantic Feature Aggregation for Gender Identification in Russian Facebook
1 Introduction
2 Related Work
2.1 Feature Aggregation for Author Profiling in Social Media
2.2 Topic Modelling
2.3 Distributional Clustering
3 Dataset
4 Feature Aggregation Models
4.1 LDA
4.2 Author-Topic Model
4.3 Distributional Clustering
4.4 Automatic Label Assignment
5 Author Gender Profiling
5.1 Experiment
5.2 Results
5.3 Correlation Analysis
6 Conclusions
References
Using Linguistic Activity in Social Networks to Predict and Interpret Dark Psychological Traits
1 Introduction
2 Method
2.1 Psychometrics
2.2 Topic Models
2.3 Predictive Models
2.4 Statistical Analysis
3 Experiment
3.1 Data Collection
3.2 Data Preprocessing
3.3 Implementation Details
4 Results
4.1 Prediction
4.2 Statistical Analysis
5 Discussion
6 Conclusion
References
Boosting a Rule-Based Chatbot Using Statistics and User Satisfaction Ratings
1 Introduction
2 Related Work
3 Overview of the Rule-Based Chatbot
4 Raw Data and Task Definition
4.1 Data
4.2 Approach Chosen
5 Data Preparation
6 Experimental Setup
6.1 Data Preprocessing
6.2 Baselines
6.3 Proposed Systems
7 Results and Discussion
7.1 System Performance
7.2 Difficulty of the Task
8 Conclusion and Future Work
References
Speech Processing
Deep Learning for Acoustic Addressee Detection in Spoken Dialogue Systems
Abstract
1 Introduction
2 Related Work
3 Experimental Data
4 Features
5 Models
6 Results
7 Conclusions
Acknowledgments
References
Deep Neural Networks in Russian Speech Recognition
1 Introduction
2 Related Work
3 Architectures of Neural Networks for Acoustic Modeling
3.1 LSTM
3.2 CNN
3.3 ResNet
3.4 RCNN
4 Datasets
4.1 Dataset for the Acoustic Models
4.2 Dataset for the Language Model
5 Speech Recognition System Implementation
6 Experiments and Results
6.1 Baseline
6.2 MLP
6.3 LSTM
6.4 CNN
6.5 ResNet
6.6 RCNN
6.7 Comparing of Models
6.8 New Model
6.9 Summarization
7 Conclusion
References
Combined Feature Representation for Emotion Classification from Russian Speech
Abstract
1 Introduction and Related Work
2 Proposed Method
2.1 Feature Extraction
2.2 Feature Representation
2.3 Classification
3 Experimental Settings and Results
3.1 RUSLANA Database
3.2 Experimental Results
4 Conclusion
Acknowledgments
References
Information Extraction
Active Learning with Adaptive Density Weighted Sampling for Information Extraction from Scientific Papers
1 Introduction
2 Related Work
3 Sampling Strategies for Active Learning
4 Task-Independent Features and Classification Pipeline
5 Annotation Tool
6 Experiments
6.1 Data
6.2 Evaluation Without Active Learning
6.3 Evaluation with Active Learning
6.4 Corpus Improvement with Active Learning
7 Conclusion
References
Application of a Hybrid Bi-LSTM-CRF Model to the Task of Russian Named Entity Recognition
1 Introduction
2 Neuronal NER Models
2.1 Long Short-Term Memory Recurrent Neural Networks
2.2 Bi-LSTM
2.3 CRF Model for NER Task
2.4 Combined Bi-LSTM and CRF Model
2.5 Neuro NER Extensions
3 Experiments
3.1 Datasets
3.2 External Word Embedding
3.3 Results
4 Discussion
5 Conclusions
References
Web-Scale Data Processing
Employing Wikipedia Data for Coreference Resolution in Russian
Abstract
1 Introduction
2 Related Work on Topic
3 Using Semantic Features from Wikipedia Data to Improve Results of Coreference Resolution
3.1 Text Preprocessing and Feature Extraction
3.2 Adding Wikipedia Data
4 Results
4.1 Discussion
4.2 Future Work
References
Building Wordnet for Russian Language from Ru.Wiktionary
1 Introduction
2 Related Work
3 Data
4 Algorithm Description
4.1 Synonym Relations Extraction
4.2 Hierarchical Links Extraction
4.3 Links Cleaning
5 Results
6 Conclusion and Future Work
References
Corpus of Syntactic Co-Occurrences: A Delayed Promise
Abstract
1 Online Resources on Word Combinations in Russian
2 Method and Used Corpora
3 CoSyCo Database and Site
4 Evaluation
5 Conclusion
References
Computation Morphology and Word Embeddings
A Close Look at Russian Morphological Parsers: Which One Is the Best?
Abstract
1 Introduction
2 Previous Work
3 Russian Morphological Parsers
3.1 Mystem
3.2 pymorhpy2
3.3 TreeTagger
3.4 FreeLing
4 Methodology
4.1 Corpora
4.2 POS Tagsets and Verb Lemmas
4.3 Evaluation Measures
5 Results and Discussion
5.1 Lemmatization
5.2 POS Tagging
6 Conclusion
Acknowledgments
References
Morpheme Level Word Embedding
1 Introduction
2 Related Work
3 Vocabularies
4 Algorithm for Segmentation a Word to Morphemes
4.1 Learning
4.2 Segmentation
5 Morpheme Embedding
6 Word Embedding Correction
7 Experiments
8 Conclusions and Future Work
References
Comparison of Vector Space Representations of Documents for the Task of Information Retrieval of Massive Open Online Courses
1 Introduction
2 Related Work
3 Approach to Comparison of Vector Representations
3.1 Corpus
3.2 Preprocessing
3.3 Vector Space Models
3.4 Processing Query
3.5 Human Judgment
3.6 Evaluation Metrics
4 Results and Discussion
5 Conclusion
References
Machine Learning
Interpretable Probabilistic Embeddings: Bridging the Gap Between Topic Models and Neural Networks
1 Introduction
2 Related Work
3 Probabilistic Word Embeddings
4 Additive Regularization and Embeddings for Multiple Modalities
5 Experiments
6 Conclusions
References
Multi-objective Topic Modeling for Exploratory Search in Tech News
1 Introduction
2 Probabilistic Topic Modeling and Additive Regularization
3 Topic-Based Exploratory Search
4 Experiments with Topic-Based Search
5 Model Parameters Optimization
6 Conclusions
References
A Deep Forest for Transductive Transfer Learning by Using a Consensus Measure
1 Introduction
2 Deep Forest
3 Consensus Measures and Training the TLDF
3.1 Weighted Average of Class Probabilities
3.2 The Shannon Entropy as a Consensus Measure
4 Convex Measure of the Transfer Learning Consistence
5 An Algorithm for the TLDF Training
6 Numerical Experiments
7 Conclusion
References
Russian Paraphrase Detection Shared Task
ParaPhraser: Russian Paraphrase Corpus and Shared Task
Abstract
1 Introduction
2 Background
2.1 Paraphrase Extraction and Recognition
2.2 Paraphrase Corpora
3 The ParaPhraser Project
3.1 The Construction Process
3.2 Crowdsourcing
3.3 Evaluation
4 Shared Task
4.1 The Task
4.2 Baselines
4.3 Results
4.4 Experiments
5 Conclusion
References
Effect of Semantic Parsing Depth on the Identification of Paraphrases in Russian Texts
Abstract
1 Introduction
2 Parser
2.1 The SemSin System
2.2 Syntactic Parsing and Semantic Analysis Using SemSin
3 Text Analysis
3.1 Lemmatization
3.2 Semantics. Accounting for Classes
3.3 Semantics. Synonymy and Additional Classes
3.4 Dependency Tree
4 Results
5 Conclusion
References
RuThes Thesaurus in Detecting Russian Paraphrases
1 Introduction
2 Related Work
3 RuThes Thesaurus
4 Using RuThes Synonyms in News Article Clustering
5 RuThes in Russian Paraphrasing Task
5.1 Russian Paraphrasing Task
5.2 Evaluating Thesaurus-Based Features in Paraphrase Detection
5.3 Finding the Best Thesaurus Feature
5.4 Combining Thesaurus Features with Other Features
6 Conclusion
References
Knowledge-lean Paraphrase Identification Using Character-Based Features
Abstract
1 Introduction
2 Related Work
3 Paraphrase Corpora
3.1 The Microsoft Paraphrase Corpus
3.2 The Plagiarism Detection Corpus
3.3 The Twitter Paraphrase Corpus
3.4 A Turkish Paraphrase Corpus
3.5 A Russian Paraphrase Corpus
4 Knowledge-Lean Paraphrase Identification
4.1 Representing Paraphrase Pairs
4.2 Classifier Training
4.3 Feature Scaling
4.4 Experiments
5 Combination of Word- and Character-Based Features
6 The Russian Paraphrase Task
6.1 Three Class Versus Binary Classification
6.2 Results
7 Discussion and Conclusions
Acknowledgements
References
Paraphrase Detection Using Machine Translation and Textual Similarity Algorithms
1 Introduction
1.1 Motivation
1.2 Objective
1.3 Task Description
2 Related Work
3 Data Set
4 Baseline
4.1 Algorithm
4.2 Results
5 Algorithm
5.1 Brief Explanation
5.2 Detailed Description
5.3 Feature Vector Structure for Each One of the Three Translations
6 Comparison of Toolkits on First Task (3-Way Classification)
6.1 Results
6.2 Confusion Matrix
7 Ablation Test and Its Analysis on Second Task (2-Way Classification)
7.1 Results
7.2 Confusion Matrix
7.3 Identifying Best SEMILAR Toolkit Score
8 Comparison of Translation Engines for Second Task (2-Way Classification)
9 Error Analysis
9.1 False Positive: Mistakenly Predicted as `Paraphrase'
9.2 False Negatives: Mistakenly Predicted as `Non-paraphrase'
10 Conclusions and Future Work
11 Repository
A Appendix: Algorithm Data-Flow
B Appendix: Ablation Test Tables
References
Character-Level Convolutional Neural Network for Paraphrase Detection and Other Experiments
1 Introduction
2 Dataset and Results of Participation
3 Convolutional Neural Network
3.1 Character Embeddings
3.2 Pre-trained Word Embeddings
4 Rule-Based
5 Classic Machine Learning Approaches
5.1 SVM with Basic Features
5.2 SVM Classifier with Advanced Features
6 Deep Neural Network Architectures
6.1 Deep Neural Network (MayoNLP)
6.2 Deep Convolutional Neural Network
7 Discussion
8 Related Work
9 Conclusions
References
Erratum to: Multi-objective Topic Modeling for Exploratory Search in Tech News
Erratum to: Chapter "Multi-objective Topic Modeling for Exploratory Search in Tech News" in: A. Filchenkov et al. (Eds.): Artificial Intelligence and Natural Language, CCIS 789, https://doi.org/10.1007/978-3-319-71746-3_16
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken