Slot Grammar. Zahra Solgi. June 18, Universität Tübingen

Similar documents
Basic Syntax. Doug Arnold We review some basic grammatical ideas and terminology, and look at some common constructions in English.

CS 598 Natural Language Processing

Compositional Semantics

Chapter 4: Valence & Agreement CSLI Publications

Universal Grammar 2. Universal Grammar 1. Forms and functions 1. Universal Grammar 3. Conceptual and surface structure of complex clauses

Syntax Parsing 1. Grammars and parsing 2. Top-down and bottom-up parsing 3. Chart parsers 4. Bottom-up chart parsing 5. The Earley Algorithm

Introduction to HPSG. Introduction. Historical Overview. The HPSG architecture. Signature. Linguistic Objects. Descriptions.

Intra-talker Variation: Audience Design Factors Affecting Lexical Selections

Objectives. Chapter 2: The Representation of Knowledge. Expert Systems: Principles and Programming, Fourth Edition

Context Free Grammars. Many slides from Michael Collins

AQUA: An Ontology-Driven Question Answering System

BULATS A2 WORDLIST 2

ENGBG1 ENGBL1 Campus Linguistics. Meeting 2. Chapter 7 (Morphology) and chapter 9 (Syntax) Pia Sundqvist

THE VERB ARGUMENT BROWSER

Underlying and Surface Grammatical Relations in Greek consider

Inleiding Taalkunde. Docent: Paola Monachesi. Blok 4, 2001/ Syntax 2. 2 Phrases and constituent structure 2. 3 A minigrammar of Italian 3

Some Principles of Automated Natural Language Information Extraction

"f TOPIC =T COMP COMP... OBJ

Proof Theory for Syntacticians

Words come in categories

Myths, Legends, Fairytales and Novels (Writing a Letter)

Update on Soar-based language processing

Adapting Stochastic Output for Rule-Based Semantics

Parsing of part-of-speech tagged Assamese Texts

Developing a TT-MCTAG for German with an RCG-based Parser

Towards a Machine-Learning Architecture for Lexical Functional Grammar Parsing. Grzegorz Chrupa la

Enhancing Unlexicalized Parsing Performance using a Wide Coverage Lexicon, Fuzzy Tag-set Mapping, and EM-HMM-based Lexical Probabilities

Case government vs Case agreement: modelling Modern Greek case attraction phenomena in LFG

Formulaic Language and Fluency: ESL Teaching Applications

Constraining X-Bar: Theta Theory

The Discourse Anaphoric Properties of Connectives

LTAG-spinal and the Treebank

1/20 idea. We ll spend an extra hour on 1/21. based on assigned readings. so you ll be ready to discuss them in class

AN EXPERIMENTAL APPROACH TO NEW AND OLD INFORMATION IN TURKISH LOCATIVES AND EXISTENTIALS

The building blocks of HPSG grammars. Head-Driven Phrase Structure Grammar (HPSG) HPSG grammars from a linguistic perspective

5 th Grade Language Arts Curriculum Map

Character Stream Parsing of Mixed-lingual Text

A Framework for Customizable Generation of Hypertext Presentations

Specifying a shallow grammatical for parsing purposes

Natural Language Processing. George Konidaris

A Minimalist Approach to Code-Switching. In the field of linguistics, the topic of bilingualism is a broad one. There are many

Using dialogue context to improve parsing performance in dialogue systems

Construction Grammar. University of Jena.

Advanced Grammar in Use

Derivational: Inflectional: In a fit of rage the soldiers attacked them both that week, but lost the fight.

Project in the framework of the AIM-WEST project Annotation of MWEs for translation

Modeling Attachment Decisions with a Probabilistic Parser: The Case of Head Final Structures

MODELING DEPENDENCY GRAMMAR WITH RESTRICTED CONSTRAINTS. Ingo Schröder Wolfgang Menzel Kilian Foth Michael Schulz * Résumé - Abstract

SEMAFOR: Frame Argument Resolution with Log-Linear Models

Leveraging Sentiment to Compute Word Similarity

The Role of the Head in the Interpretation of English Deverbal Compounds

Reading Grammar Section and Lesson Writing Chapter and Lesson Identify a purpose for reading W1-LO; W2- LO; W3- LO; W4- LO; W5-

Interfacing Phonology with LFG

The Interface between Phrasal and Functional Constraints

arxiv:cmp-lg/ v1 16 Aug 1996

Ch VI- SENTENCE PATTERNS.

The presence of interpretable but ungrammatical sentences corresponds to mismatches between interpretive and productive parsing.

11/29/2010. Statistical Parsing. Statistical Parsing. Simple PCFG for ATIS English. Syntactic Disambiguation

An Interactive Intelligent Language Tutor Over The Internet

Building an HPSG-based Indonesian Resource Grammar (INDRA)

Accurate Unlexicalized Parsing for Modern Hebrew

Basic Parsing with Context-Free Grammars. Some slides adapted from Julia Hirschberg and Dan Jurafsky 1

Word Stress and Intonation: Introduction

Chunk Parsing for Base Noun Phrases using Regular Expressions. Let s first let the variable s0 be the sentence tree of the first sentence.

Linking Task: Identifying authors and book titles in verbose queries

Loughton School s curriculum evening. 28 th February 2017

L1 and L2 acquisition. Holger Diessel

Pseudo-Passives as Adjectival Passives

Approaches to control phenomena handout Obligatory control and morphological case: Icelandic and Basque

A Computational Evaluation of Case-Assignment Algorithms

Taught Throughout the Year Foundational Skills Reading Writing Language RF.1.2 Demonstrate understanding of spoken words,

BASIC ENGLISH. Book GRAMMAR

Emmaus Lutheran School English Language Arts Curriculum

Grammars & Parsing, Part 1:

Subject: Opening the American West. What are you teaching? Explorations of Lewis and Clark

Can Human Verb Associations help identify Salient Features for Semantic Verb Classification?

THE INTERNATIONAL JOURNAL OF HUMANITIES & SOCIAL STUDIES

Feature-Based Grammar

Agnès Tutin and Olivier Kraif Univ. Grenoble Alpes, LIDILEM CS Grenoble cedex 9, France

(12) United States Patent Bernth et al.

Minimalism is the name of the predominant approach in generative linguistics today. It was first

Specifying Logic Programs in Controlled Natural Language

Prediction of Maximal Projection for Semantic Role Labeling

Target Language Preposition Selection an Experiment with Transformation-Based Learning and Aligned Bilingual Data

What Can Neural Networks Teach us about Language? Graham Neubig a2-dlearn 11/18/2017

The Structure of Multiple Complements to V

Dear Teacher: Welcome to Reading Rods! Reading Rods offer many outstanding features! Read on to discover how to put Reading Rods to work today!

Graph Alignment for Semi-Supervised Semantic Role Labeling

ON THE SYNTAX AND SEMANTICS

A First-Pass Approach for Evaluating Machine Translation Systems

Houghton Mifflin Reading Correlation to the Common Core Standards for English Language Arts (Grade1)

LNGT0101 Introduction to Linguistics

SAMPLE. Chapter 1: Background. A. Basic Introduction. B. Why It s Important to Teach/Learn Grammar in the First Place

Procedia - Social and Behavioral Sciences 154 ( 2014 )

Chapter 9 Banked gap-filling

LING 329 : MORPHOLOGY

Control and Boundedness

Dickinson ISD ELAR Year at a Glance 3rd Grade- 1st Nine Weeks

Semi-supervised methods of text processing, and an application to medical concept extraction. Yacine Jernite Text-as-Data series September 17.

An Introduction to the Minimalist Program

Transcription:

1 Slot Grammar Zahra Solgi Universität Tübingen June 18, 2016

2 Slot Grammar Overview Slot Grammar? what is the use of that?

3 Slot Grammar Overview Slot Grammar? what is the use of that?

Slot Grammar Overview Slot Grammar? what is the use of that? what are ESG,FSG,SSG,ISG,BPSG,GSG?

Slot Grammar SG: provides a convenient means for writing practical, broad-coverage grammars for natural language extra eplantaion as aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaa aaaaaaaaaaaaaaaaaa the reduction of differences between grammars of different languages

Slot Grammar Overview SG: provides a convenient means for writing practical, broad-coverage grammars for natural language An SG is a Dependency tree which reduces the differences between grammars of different languages

7 Using Slot Grammar overviw the input is segmented by a "Sentences" separator the output of SG analysis is a pars tree

8 Using Slot Grammar overview comparison between Dependency grammar and constituency grammar

9 Using Slot Grammar overview Dependency grammar vs. constituency grammar

10 Deep parsing Surface structure is determined by modifier structure Deep structure is determined by logical argument frames

Using Slot Grammar overview subj subject obj direct object iobj indirect object comp predicate complement objprep object of preposition ndet NP determiner : :

SG ESG parse tree: Chandeliers look great but nowadays do not usually use these items from which their name is derived. 12

13 Using Slot Grammar Complement slots determined by the properties of their headword (e.g verb which take subjects and objects) Adjunct slots determined by the part of speech of their headword (e.g. verbs can take an adverb)

14 parse Nodes Headword ID Logical argument frame Features Modifier structure

15 SG lexical entries Morpholexical Analysis: how to determine frames? Look up word in the provided SG Lexicon and match its use in context to a sense frame specified in the Lexicon Lexical entries Part of speech e.g. noun, verb, adjective, etc. Complement slot frame Features syntactic features or semantic types e.g. object, property, event, living being Numerical score rate sense frames Subject area e.g. computers, medicine Support verb construction

IBM imrovment on SG lexicon Match noun frames with verb frames E.g. encode a relationship between celebration and celebrate Helps match questions to answers Augmentation of ESG base lexicon (using WordNet) Increases number of entries Indicates semantic types Noun-verb correspondences E.g. verb defer has indicated noun-forms deferral, deference, etc. Chunk Lexicons Handle multiword entries (e.g. Sing a Song of Sixpence ). LAT Reward Features Aid in identification of answer types

syntactic analysis 1. Combine tokens into syntactic constituents 2. Bottom-up, left-right organization of constituents into slots 3. Subtrees build phrases 4. Phrases are scored according to lexical use of constituents, rules in grammar

Predicate argument Simplifies and generalizes result of ESG parse 1. Elements change exact semantic meaning but in general are not essential to its core meaning. 2. Does not process original text. Instead modifies the output of the ESG parse. I heard that Edison invented the phonograph in 1877. I heard that Edison invented a phonograph in 1877. I heard Edison invented the phonograph in 1877. I heard that Edison was inventing the phonograph in 1877. I heard that the phonograph was invented by Edison in 1877. E.g, Have different meanings generate different ESG parse trees 1. Exact semantic meaning irrelevant since they all contain the same evidence to answer a question Who invented the phonograph?

19 slot filling analysis Slot filling for Mary gave John a book give (e, x, y, z) means e is an event where x gives y to z

20 Logical Representation analysis Slot filling for Mary gave a book to John

21 Ingredients Ingredients of Slot Grammar analysis Structure

Extrsposition phrase (X, H, Sense, Features, SlotFrame, Ext, Modes) the argument Ext is used to hold Extraposed slots i.e: slots that can be filled by left-extraposed phrases loke who in whodidalicetrytofind. The list Ext consists of internal form and one element Ingerients of shells: a declaration of extraposer slots extraposed filler rules Extraposer slots are slots that allow extrapositions out of their fillers

Coordination The following method for handling coordination was outlined in (MacCord 1980) then was implemented in the recent adaptation of Slot Grammar to logic programing The form of coordination in a phrase: LM Preconj LC Conj RC RM

Coordination The subsrings indecated are intrepretated as follows: Conj cooredinationg conjunction (like and or or) or a punctuation like comma Preconj optional preconjunction that can accompany Conj (both for and) LC and RC left and right conjuncts respectively. (one single phrase) LM and RM (optional) left and right common modifier, respectively. (several phrases)

25 Coordination Let s make an Example: The man LM sees LC and Conj probably hears RC the car. RM John sees LC and Conj Mary hears RC the car. RM

More and More We have given an overview of ESG analysis and indicated the central role of slots and slot frames. conclusion and Future Work 1. expansion of the semantic type system and its use in parsing; 2. incorporation of word-sense disambiguation, probably with senses of less granularity than in WordNet. 3. indication in parse trees of scoping of generalized quantifiers and focusing adverbs, etc. 4. development of specialized lexicons and methods for handling very large lexicons 5. continued improvement of coverage of SG via regression testing.