Human Language Technology. Challenges of the Information Society

This book constitutes the refereed proceedings of the Third Language and Technology Conference: Human Language Technologies, LTC 2007, held in Poznan, Poland in October 5-7, 2007.

The 40 revised papers presented were carefully reviewed and selected. The papers are organized in topical sections on speech processing, computational morphology, parsing and generation, computational semantics, digital language resources, WordNet, information retrieval/extraction and machine translation.

Klappentext

Half a centuryago not manypeople had realizedthat a new epoch in the history of homo sapiens had just started. The term "Information Society Age" seems an appropriate name for this epoch. Communication was without a doubt a lever of the conquest of the human race over the rest of the animate world. There is little doubt that the human racebegan when our predecessorsstarted to communicate with each other using language.This highly abstractmeans of communicationwas probably one of the major factors contributing to the evolutionary success of the human race within the animal world. Physically weak and imperfect, humans started to dominate the rest of the world through the creation of communication-based societies where individuals communicated initially to satisfy immediate needs, and then to create, accumulate and process knowledge for future use. The crucial step in the history of humanity was the invention of writing. It is worth noting that writing is a human invention, not a phenomenon resulting from natural evolution. Humans invented writing as a technique for recording speech as well as for storing and facilitating the dissemination of knowledge across the world. Humans continue to be born illiterate, and therefore teaching and conscious supervised learning is necessary to maintain this basic social skill.

Inhalt

Speech Processing.- Exploring the Structure of Broadcast News for Topic Segmentation.- Application of Slope Filtering to Robust Spectral Envelope Extraction for Speech/Speaker Recognition.- Spoken Language Interface for Mobile Devices.- A Study on Bilingual Speech Recognition Involving a Minority Language.- Annotated Corpus of Polish Spoken Dialogues.- Triphone Statistics for Polish Language.- Computational Morphology.- ROG - A Paradigmatic Morphological Generator for Romanian.- Building a Morphosyntactic Lexicon and a Pre-syntactic Processing Chain for Polish.- A Relational Model of Polish Inflection in Grammatical Dictionary of Polish.- Parsing and Generation.- Syntactic Spreadsheets: In Search for a Human-Readable Representation of Parse Tree Forests.- A Hybrid System for Named Entity Metonymy Resolution.- Spejd: A Shallow Processing and Morphological Disambiguation Tool.- Flexible Natural Language Generation in Multiple Contexts.- A Measure of the Number of True Analogies between Chunks in Japanese.- Shallow Parsing of Transcribed Speech of Estonian and Disfluency Detection.- Mining Parsing Results for Lexical Correction: Toward a Complete Correction Process of Wide-Coverage Lexicons.- Efficient Parsing Using Recursive Transition Networks with Output.- The Deep Parser for Polish.- Computational Semantics.- Automatically Determining Attitude Type and Force for Sentiment Analysis.- A Readability Checker Based on Deep Semantic Indicators.- The DANTE Temporal Expression Tagger.- Hybrid Unsupervised Learning to Uncover Discourse Structure.- Pronoun Resolution in Turkish Using Decision Tree and Rule-Based Learning Algorithms.- Digital Language Resources.- LMF-QL: A Graphical Tool to Query LMF Databases for NLP and Editorial Use.- Novelty Extraction from Special and Parallel Corpora.- Construction of Text Corpus of Polish Using the Internet.- A Predicate Database for Assisting the Design of a Lexicon-Grammar of Predicative Nouns.- A Standard Lexical-Terminological Resource for the Bio Domain.- Extracting Collocations in Contexts.- WordNet.- Putting Semantics into WordNet's "Morphosemantic" Links.- Leveraging Parallel Corpora and Existing Wordnets for Automatic Construction of the Slovene Wordnet.- An Algorithm for Building Lexical Semantic Network and Its Application to PolNet - Polish WordNet Project.- Information Retrieval/Extraction.- ECODE: A Definition Extraction System.- Using Graph-Based Indexing to Identify Subject-Shift in Topic Tracking.- Result Aggregation for Knowledge-Intensive Multicultural Name Matching.- Comparison of String Distance Metrics for Lemmatisation of Named Entities in Polish.- An Iterative Model for Discovering Person Coreferences Using Name Frequency Estimates.- Machine Translation.- Hapax Legomena: Their Contribution in Number and Efficiency to Word Alignment.- Statistical Machine Translation from Slovenian to English Using Reduced Morphology.- Semi-automatic Creation of a Dictionary of Nominal Compounds.