Courses' slots:
Week two 9:00 - 10:30
Language and Logic introductory course:
Topics in the Semantics of Interrogative Clauses.
Teachers
Abstract: The purpose of the class is to present in a systematic way some of the most influential lines of investigations pertaining to the semantics of questions. We will start by presenting two related types of theories, namely theories based on "sets of answers" (Hamblin 1973, Karttunen 1977), on the one hand, and theories based on "partition semantics", on the other hand (Gronendijk & Stockhof 1982, 1984), and discuss their strengths and weaknesses. This will lead us to an extensive discussion of embedded interrogatives (including topics such as weak and strong exhaustivity in relation to NPI licensing, the distinction between de dicto and de re readings, extensional vs. intensional question embedding predicates, quantificational variability). We will provide a compositional account of the meaning of wh-questions, which will allow us to address more specific topics such as identity questions, functional and pair-list readings of wh-questions, alternative questions.
Language and Computation advanced course:
Psycho-computational issues in Morphology Learning and Processing.
Teacher
- Vito Pirrelli ()
By providing a comprehensive overview of current machine-learning, psycholinguistic and theoretical linguistic literature on the topic, the course is intended to answer the following questions. How are words singled out of their embedding input stream? How are they processed and eventually understood in working memory? Are morphologically complex words stored in long-term memory as a whole or are they rather composed "on-line" in working memory from sub-lexical constituents? Do formal regularity and morpho-semantic transparency play any role in this? Does word-level knowledge require parallel development of form and meaning representations, or do the latter develop independently at a different pace to interact only at later stages? To what extent does past knowledge affect on-line word processing? What principles govern this knowledge? Are they morphology-specific or are they rather based on brain memory structures generically devoted to the ordered activation of items in time? Do they capture local, syntagmatic relations among sub-lexical co-occurring constituents, or also enforce more global paradigmatic constraints over classes of such constituents in complementary distribution?
Language and Logic foundational course:
Meaning Composition: Empricial Problems and Formal Solutions.
Teacher
Abstract:
This course provides an overview two of the main empirical
problems that have emerged in the development of models for
meaning composition in natural language, the tradeoffs that are
involved in solving these problems, and some of the different
techniques that have been proposed as solutions. The goal is
twofold: to make students with logic backgrounds aware of the
reasons why the composition of natural language meanings is not
a trivial problem (even though at some levels it might seem that
way), and to familiarize students with linguistics backgrounds
with some of the main alternative techniques for meaning
composition, their similarities and differences, and their pros
and cons. The course will presuppose only a minimal familiarity
with basic grammatical concepts and predicate logic.
An elegant theory of meaning composition for natural language
might be expected to meet the following desiderata, among
others:
-It should respect independently-motivated results of research
on morphology, syntax, and the lexicon.
- It should be grounded in an independently motivated theory of
what lexical meanings are like.
- It should avoid idiosyncratic composition rules to the extent
possible.
- It should be expressible in a sound and computationally
tractable logic.
However, natural language data sometimes make a maximally
elegant theory difficult. Perhaps the best-studied problem for
the meaning composition in this respect has been
quantification. In this course, we will focus on two additional
problems which have driven various kinds of alternative meaning
composition strategies: bare nominals and incorporation on the
one hand, and so-called "intersective"
vs. "nonintersective" modification, on the other. We will
develop a sense of the general nature of the problems these
phenomena pose, as well as a global vision of the issues the
proposed solutions raise.
The plan for the course is the following:
Day 1: The basics: Classic "rule-to-rule"
vs. "shake-and-bake" approaches to
composition. [Discussion of work by Bach, Carpenter, Dowty,
Klein & Sag, Montague, and others]
Days 2-3: The empirical problem: Bare nominals and
incorporation. The solutions: type shifting, the separation of
syntactic and semantic saturation, Discourse Representation
Theory-based alternatives. [Discussion of work by Chung &
Ladusaw, Dayal, de Hoop, Espinal & McNally, Farkas & de Swart,
Kamp, Partee, Van Geenhoven, and others]
Days 4-5: The empirical problem: Intersective
vs. nonintersective modification. The solutions: type coercion,
enriched lexical representations, ad-hoc composition
rules. [Discussion of work by Asher, Larson, McNally, Montague,
Pustejovsky, and others]
Language and Computation introductory course:
Standard XML query languages for natural language processing.
Teacher
Course material: u_schaefer_xml_query.pdf
Abstract:This course will introduce three standard XML query languages that have been designed by the World Wide Web Consortium (W3C), XPath, XSLT and XQuery. Although various query languages have been proposed and developed for accessing annotated corpora, they are often tailored to specific formats and phenomena. This course will focus on the standard query languages for which multiple and very efficient implementations exist that run on almost any platform. Applications and examples are presented not only for corpus access, but also other NLP-related tasks such as accessing RDF ontologies and integrating NLP component output. Finally, the course will also briefly show the frameworks that are used to embed the query languages in popular programming languages.
Logic and Computation advanced course:
Ontologies: Structuring, Modularity, and Heterogeneity.
Teachers
- Stefano Borgo ()
- Oliver Kutz ()
The design of formal ontologies is an interdisciplinary area of
research that draws on logic, philosophy, cognitive science,
linguistics, as well as computer science, with major
applications in the Semantic Web. As the scope and relevance of
ontologies grows, both for supporting Semantic Web applications
and for knowledge-rich processing in general, the issue of
re-using/importing developed ontological components takes on an
ever more critical role. The current solutions being pursued
within OWL-oriented Semantic Web approaches have some severe
limitations in this respect. For the next generation of
ontology-based systems, it will be essential to move beyond
this.
To achieve this, we present major methodologies and techniques
to correctly construct, modify, and relate ontologies -
understood in a broad sense as logical theories formulated in
various formal languages - with an emphasis on heterogeneity,
structuring and modularity, as well as foundations of ontology
design. As illustrative examples, we will discuss prominent
ontologies from the spatial, philosophical and linguistic
domains. These will be analysed and structured using the Common
Algebraic Specification Language (CASL), and shown 'at
work' employing the tool HeTS, offering (heterogeneous)
reasoning support for structured ontologies and providing
powerful new mechanisms for reusing ontological components or
modules. A Live-CD for hands-on experimentation with HeTS will
be distributed to all participants.
Logic and Computation foundational course:
Logics of Rational Agency.
Teacher
- Eric Pacuit ()
Course material: lori-notes.pdf
Abstract:
Thinking about rational agents interacting over time is at the
center of many research communities represented at ESSLLI. This
course will introduce the main research themes and conceptual
issues surrounding rational agency. The primary objective is to
understand the complex phenomena that arise when rational
agents interact and how to incorporate these phenomena into
formal models. Studying rational agents involves many different
aspects including (but not limited to) action, knowledge,
belief, desires, and revision. This course covers all these
ingredients toward the goal of understanding how these things
work together. Specific topics that will be introduced during
the course include 1. logics of knowledge and belief,
2. information dynamics and belief revision, 3. logics of
preference and preference change, 4. logics of motivational
mental attitudes, and 5. logics of individual and collective
action and 6. group phenomena and issues of social choice. In
fact, not all parts of this story have been developed within
one single discipline. The course will also bring together
several research programs: from philosophy, computer science,
logic, and game theory, and try to see their various
contributions in one coherent manner.
http://ai.stanford.edu/~epacuit/classes/esslli/log-ratagency.html
Week two 11:00 - 12:30
Logic and Computation advanced course:
Logics with Counting.
Teacher
Course material: pratt-hartmann.pdf
Abstract:This course presents a survey of decidable fragments of first-order logic with counting quantifiers. Such fragments include: (i) the numerical syllogistic, (ii) the 1-variable fragment with counting, (iii) graded modal logic, (iv) the guarded two-variable fragment with counting and (v) the two-variable fragment with counting. Such logics present special difficulties for the student of computational logic, because their complexity-theoretic analysis often requires specialized techniques. This course will give a comprehensive treatment of these techniques, culminating in a proof that the satisfiability and finite satisfiability problems for the two-variable fragment with counting quantifiers are both NEXPTIME-complete.
Language and Logic introductory course:
Event Semantics and Adverbial Modification.
Teachers
Abstract: The course gives a general overview of (Neo-)Davidsonian event semantics and its motivation from adverbial modification. It furthermore introduces the notion of event structure, both from a conceptual and a model-theoretic point of view. An important issue concerns how far event structure, aspectual structure and argument structure are mutually related and whether it is possible or even preferable to reduce them to one. To identify elements of the event structure, argument structure and/or aspectual structure, we focus on the (un-)availability of various adverbial modifiers, in particular manner, temporal, aspectual and spatial ones, as well as the relation of such modifiers to the overall structure, e.g. the issue of high (event-external) vs. low (event-internal) adverbs. The presentation of the cross-linguistic diversity in the marking of voice and argument and adjunct roles, and phenomena like serial verb constructions, lead to a general discussion of the nat ure of the verbal category.
Language and Logic advanced course:
Proof-Theoretic Semantics.
Teacher
Course material: proof_theoretic_semantics.html
Abstract:
Proof-Theoretic Semantics (PTS) is an alternative to
model-theoretic (or truth-condition) semantics. It is based on
the idea that the central notion in terms of which meanings are
assigned to expressions is that of proof rather than truth. In
this sense PTS is inferential rather than denotational in
spirit. Although the claim that meaning is use has been quite
prominent in philosophy for more than half a century, the
model-theoretic approach has always dominated formal
semantics. However, within general proof theory several formal
approaches to PTS have been developed which promise to provide
an alternative to the model-theoretic approach. After recalling
certain basics from proof theory of natural deduction, this
tutorial presents traditional approaches to PTS in the spirit of
Dummett, Prawitz and Martin-Löf as well as advanced conceptions
based on "definitional reasoning" based on work of the
author.
http://www-ls.informatik.uni-tuebingen.de/esslli/
Language and Computation foundational course:
Case, Scrambling and Default Word Order.
Teachers
- Miriam Butt ()
- Heike Zinsmeister ()
Course material: 01-CaseScramblingWordOrder_reader.pdf 02-Mueller1999.pdf 03-Mueller2002.pdf 04-BresnanEtAL2007.pdf 05-Evert2006.pdf 06-LuedelingEvertBaroni2007.pdf 07-Meurers2005.pdf 08-BaderHaeussler.pdf 09-CahillForstRohrer2007.pdf 10-FilippovaStrube2007.pdf 11-Forst2007.pdf 12-Keller2000.pdf 13-PatilEtAl2008.pdf 14-SchulteImWalde2002.pdf
Abstract:
Many of the world's languages are so-called "free word
order" languages, whereby the major arguments of a clause
can be scrambled quite freely. This scrambling generally goes
hand-in-hand with a robust case marking system and some means of
verb-argument agreement (usually verb-subject agreement, but not
always), which allows the identification of the various
arguments of the clause (i.e., which is the agent, the patient,
the goal, etc.). Sometimes, however, the correct identification
of which syntactic argument encodes which of the semantic
participants of a verb/clause can only be achieved by world or
contextual knowledge. Additionally, effects of so-called word
order freezing can be observed, whereby suddenly the word order
is not free, but is fixed if one wants a certain mapping of
semantic participants to syntactic arguments. Finally, one
generally also refers to a "default word order"
exhibited by languages which in principle allow for the
(more-or-less) free scrambling of syntactic arguments. The
theoretical status of this default word order is not clear and
this course will examine the topic of argument scrambling, word
order freezing and default word order with respect to two main
perspectives: 1) theoretical linguistics; 2) corpus linguistics.
With respect to the theoretical perspective, students will be
introduced to current theories of case and word order so that we
can examine what (if anything) these theories have to say about
default word order and word order freezing in particular. With
respect to the computational perspective, we will examine to
what degree information from corpora can help guide the analysis
and help us understand why things are scrambled when they are
scrambled and what status the "default word order"
actually has in terms of frequency and distribution. As part of
the course (one day), we will also present psycholinguistic
studies that identify (combinations of) features that determine
word order preferences.
Language and Computation introductory course:
Computational Lexical Semantics.
Teachers
- Gemma Boleda ()
- Stefan Evert ()
This course will provide students with an overview of current research in Computational Lexical Semantics, and with the necessary theoretical and methodological background to carry out their own research. Students will have an opportunity to work on practical examples, learning to tackle the difficulties mentioned above. Special emphasis will be put on the feedback between computational approaches and semantic theory.
Logic and Computation advanced course:
Reasoning with Probabilities.
Teachers
- Eric Pacuit ()
- Joshua Sack ()
Course material: epprob-outline.pdf
Abstract:
Both logic and probability provide powerful tools for reasoning
about uncertainty in a dynamic environment. Our goal in this
course is to examine logical frameworks that incorporate
probabilistic modeling of multiagent uncertainty. We will then
see how merging these two perspectives on uncertainty can help
clarify various conceptual issues and puzzles (such as the Monty
Hall puzzle or the sleeping beauty problem). The primary
objective is to explore the formal tools used by logicians,
computer scientists, philosophers and game theorists for
modeling uncertainty. We will focus on both the important
conceptual issues (eg., Dutch book arguments, updating with
probability zero events and higher-order probabilities) and the
main technical results (eg., completeness and decidability of
probabilistic modal logics).
Course Outline: The course will cover all of the following
topics:
0. Background in dynamic epistemic logic and probabilistic
models of uncertainty (including issues of measurability and
Dutch book arguments)
1. Various puzzles about probability, beliefs and time
(including the Monty Hall puzzle, the Sleeping Beauty problem
and the Absent-Minded Driver problem)
2. Higher-order probability measures (including a
definition of "common p-belief'' and uses of the concept in game
theory)
3. Modal probabilistic logic (including proofs of completeness,
decidability, and other standard results)
4. Uncertainty in a dynamic environment (including
Bayesian/Jeffrey updates)
5. Updating with probability zero events (including
lexicographic probability distributions, conditional probability
distributions and relations with belief revision models)
6. Dynamic Epistemic Probabilistic Logic (including discussion
of reduction axioms and temporal extensions)
Prerequisites: This is an advanced but self-contained
course. Students will be expected to have had some exposure to
(dynamic) epistemic logic and probabilistic logic. In
particular, it will be assumed that students have already been
introduced to epistemic logic and some of its dynamic extensions
(i.e., public announcement logic); and although we will
introduce many basic concepts of probabilistic theory (e.g.,
measure spaces), it will be expected that students have had
previous exposure to probabilistic models of uncertainty.
http://ai.stanford.edu/~epacuit/classes/esslli/epprob.html
Week two 14:00 - 15:30
Language and Logic advanced course:
Advances in Abstract Categorial Grammars: Language theory and linguistic modeling.
Teachers
Course material: esslli-acg-week-2.pdf
Abstract:The abstract categorial grammar (ACG) (de Groote 2001), a grammar formalism based on the typed lambda calculus, elegantly generalizes and unifies a variety of grammar formalisms that have been proposed for the description of formal and natural languages. The first part of this course investigates formal-language-theoretic properties of "second-order" ACGs, a subclass of ACGs that have "context-free" derivations. Their generative capacity is precisely characterized, and an efficient Earley-style algorithm is presented. The second part of the course turns to linguistic applications of ACGs and gives various illustrations of how ACGs provide flexible and explicit ways to model the syntax-semantics interface of natural language.
Language and Logic workshop:
New Directions in the Theory of Presupposition.
Organizers
Abstract:
The last ten years has seen a wealth of new developments on the topic of presupposition and, in particular, the projection problem for presupposition. While there had been considerable interest in the seventies in developing entirely pragmatic accounts of presupposition triggering and projection (Wilson, 1974, Stalnaker 1977, Grice, 1981), these accounts had generally not been sufficiently developed to match the dynamic accounts developed in the eighties in predictive power. Recent work, such as that of Schlenker (2006, 2008), however, has shown that broadly pragmatic accounts can also have considerable predictive power. In addition, trivalent approaches based on such techniques as supervaluations and the Strong Kleene connectives, which were dismissed by many long ago, have recently attracted new interest (Beaver and Krahmer, 2001, George, 2008, Fox, 2008) and have been shown capable of handling many empirical issues in presupposition projection. Thus there is no longer a clear consensus on how we should explain presupposition projection. In addition, experimental work has raised interesting questions about what the basic facts of presupposition projection are and suggests that real empirical work is needed to determine some of the subtleties (Chemla 2007). There has also been renewed interest in the triggering problem (Simons, 2001, Abusch, 2002) which naturally links up to the projection problem, as well as recent theoretical work on foundational issues such as the notion of common ground and accommodation (Beaver and Zeevat, 2004, von Fintel, 2001, 2006, Stalnaker, 2002). The purpose of this workshop is to bring together researchers on presupposition to discuss these new developments and connect some of the different theoretical and empirical questions, which are too often considered in isolation.
Language and Computation introductory course:
Intelligent Computer-Assisted Language Learning: An introduction to an emerging interdisciplinary field.
Teacher
Abstract: Intelligent Computer-Assisted Language Learning (ICALL) is a relatively young field of interdisciplinary research exploring the integration of natural language processing in foreign language teaching. The course will introduce both the theoretical issues and the practical system development aspects of ICALL and provide the student with a firm basis for understanding the current research issues. Key questions discussed include the following: Where does ICALL fit into foreign language teaching? Why are notions such as noticing and awareness from cognitive psychology important for second language acquisition and ICALL research? How can natural language processing (NLP) be adapted to process learner language? What are the challenges for NLP in detecting properties of learner language and what is know about presenting feedback to learners? What are learner models and what roles do they play in ICALL systems? And last but not least, how can shallow semantic NLP analysis be used to provide feedback on meaning in addition to feedback on form, and why is this important?
Language and Computation advanced course:
Linguistic Information Visualization.
Teachers
Course material: carpendale_penn.pdf
Abstract:
Much of what computational linguists fall back upon to improve
natural language processing and model language "understanding" is
structure that has, at best, only an indirect attestation in
observable data. The sheer complexity of these structures, and
the observable patterns on which they are based, however, usually
limits their accessibility, often even to the researchers creating
or studying them. Traditional statistical graphs and
custom-designed data illustrations fill the pages of CL papers,
providing insight into linguistic and algorithmic structures, but
visual 'externalizations' such as these are almost exclusively
used in CL for presentation and explanation.
Visualizations can also be used as an aid in the process of
research itself. There are special statistical methods, falling
under the rubric of "exploratory data analysis", and visualization
techniques just for this purpose, in fact, but these are not
widely used or even known in CL. These novel data visualization
techniques offer the potential for creating new methods that
reveal structure and detail in data. Visualization can provide new
methods for interacting with large corpora, complex linguistic
structures, and can lead to a better understanding of the states
of stochastic processes.
Instructed by a team of computational linguists and information
visualization researchers, this tutorial will bridge computational
linguistic and information visualization expertise, providing
attendees with a basis from which they can begin to leverage
information visualization in their own research. It will equip
participants with: - An understanding of the importance and
applicability of information visualization techniques to
computational linguistics research; - Knowledge of the basic
principles of information visualization theory; - The ability to
identify appropriate visualization software and techniques that
are available for immediate use and for prototyping; - A working
knowledge of research to date in the area of linguistic
visualization.
This tutorial will be an extended version of the 3-hour tutorial
offered at ACL-2008, which had 25 attendees. The instructors have
previously taught portions of the content in advanced
undergraduate and graduate courses as well. Students are expected
to have a solid background in computational linguisics. No
experience with visualization is required.
TUTORIAL OUTLINE
Day 1: Introduction; Information Visualization Theory
(representational theory, cognitive psychology, preattentive
processing, interaction & animation, assessing and validating
visualizations)
Days 2 and 4: Review of Linguistic Visualizations (document
content visualizations, text collection analysis, literary
analysis, streaming data visualization, convergence of linguistic
data and social network analysis, corpora exploration,
visualization uncertainty in statistical NLP output, linguistic
analysis, visualization of speech data)
Day 3: Tools for Visualization (software solutions: Excel,
Tableau, Spotfire, programming toolkits: prefuse, processing,
flare, InfoVis Toolkit, online tools: ManyEyes, Swivel,
collaborative visualization tools in development)
Day 5: Case Study: Visualization for Statistical MT; Open Research
Problems (CL problems that could benefit from visualization,
Visualization of language areas that need CL expertise); Closing
Logic and Computation advanced course:
Dynamic Logics for Interactive Belief Revision.
Teachers
- Alexandru Baltag ()
- Sonja Smets ()
This course is addressed to students and researchers interested in logics for reasoning about multi-agent belief revision, belief updates and knowledge updates induced by various forms of communication or interaction. It is a foundational course, designed to present to students and researchers from other fields the work done in recent years by a number of researchers on integrating ideas from Belief Revision Theory within the DEL (Dynamic-Epistemic Logic) paradigm. The course is self-contained, not assuming any background knowledge, but presupposing only some general experience or facility with using logical languages, the concepts of syntax and semantics etc. We start by presenting the main notions of "standard DEL", arguing that this logic is appropriate for updating "hard information" (unrevisable knowledge), but that it is inappropriate for "soft information" (possibly false beliefs or defeasible knowledge). We then present "belief-revision models", defining some important epistemic/doxastic notions, considering a number of logical languages for these models, and explaining the relevance of these notions to fundamental issues in contemporary Epistemology, in Computer Science and in the study of language and communication. We present various Belief Update operations and various belief-revision policies proposed by different authors, focusing on one particular proposal (the Action-Priority Update) of great generality and elegance. We present reduction laws (the "dynamic laws of Interactive Belief Revision"), complete axiomatizations, applications to dialogue games and communication strategies, connections with other research areas and open problems.
Language and Logic introductory course:
Syntax and Semantics from an Algebraic Perspective.
Teachers
- Edward Stabler ()
- Edward Keenan ()
Course material: StablerKeenan-ESSLLI09-Syllabus.pdf Keenan09-Malagasy.pdf KeenanStabler05.pdf KeenanStabler09.pdf StablerKeenan07.pdf
Abstract:
Developing perspectives from Keenan and Stabler (2003), this class will explore algebraic methods for compositionally defining and comparing various languages (syntax and semantics), with particular attention to various constructions with quantification and binding in the world's languages.
Introduction: Language structure (!=trees) and universals (!=fixed constituent orders). A universal of anaphor binding.
Evans and Levinson (BBS, forthcoming) take Chomskyan linguistics to task for insufficient recognition of the diversity of human languages. This diversity, they claim, renders universal claims few in number and largely trivial in content, whence the monolithic picture of Universal Grammar that generative grammarians present to the CogSci community at large is inaccurate and misleading. Our response to this critique exhibits a novel format of universal generalization that has the mathematical rigor desired (if not always achieved) by a Chomskyan approach, but which incorporates (morpho)syntactic diversity in an essential way. More important than diversity per se is that languages exhibit non-universal structural regularities that speakers must learn. We illustrate this with universal claims concerning anaphora -- one of the areas in which E+L criticize the Chomskyan approach
Example 1: Affixes and structure in Malagasy, without movement.
Example 2: Quantification in 'minimalist grammar', with movement. A 'directly compositional' semantics for a grammar with
'quantifier raising'.
Language similarity 1: Stable automorphisms; perspectives on probabilistic language models.
Language similarity 2: Homomorphic realizations of canonical patterns of predication, modification, and binding.
Week two 17:00 - 18:30
Logic and Computation introductory course:
Explanation in Ontology Languages.
Teachers
- Bijan Parsia ()
- Thomas Schneider ()
Course material: explanation.pdf
Abstract:This course will present current research into explanations, which can answer questions such as: "Why do certain entailments follow from my ontology?"; "Why is my ontology inconsistent?"; "Which part of my ontology do I need to modify in order to prevent an undesired entailment?" These questions are of relevance to ontology developers in various domains including bioinformatics, medicine, geography, linguistics. We will focus on proof based explanation techniques and touch on the emerging field of model based explanation.
Language and Logic advanced course:
Referent Systems.
Teachers
- Udo Klein ()
- Marcus Kracht ()
Formal systems for computing meanings often rely very strongly
on syntactic structure, and rarely on other overt clues such as
morphology. This makes them quite inflexible. The calculus of
Referent Systems, due to Kees Vermeulen and further developed
by the second lecturer, allows for much more flexible
algorithms since it accesses both syntactic and morphological
structure. The basic idea is that semantic composition is not
performed by function application, but by a semantic operation
which merges variables depending on the morphosyntactic
properties associated with them. The main burden for
calculating meanings of complex expressions is carried by the
argument structure, which interfaces syntax, morphology and
semantics. The system has been implemented (see ). The
course will not only introduce the idea of referent systems and
how they function, but also show how to implement and use the
system.
The course assumes knowledge of formal semantics (DRT, for
example) and some general knowledge of linguistics. All formal
concepts will be properly introduced.
Proposed Content:
Lec 1: Motivation and basic idea: Composing meanings
Lec 2: Referent systems
Lec 3: Case
Lec 4: Parameters
Lec 5: Agreement Morphology and its Role in Semantics
Language and Computation introductory course:
Corpus-Based Argument Structure.
Teacher
Abstract: The aim of the course is twofold. The linguistic goal is to discuss the notion of argument structure (valence) both from the syntactic and from the semantic point of view, with some emphasis on the argument/adjunct dichotomy and diathesis (argument alternations). The computational goal is to present diverse techniques of learning valence information from corpora. This automatic learning task is usually split into two stages: the linguistic stage of collecting information about the co-occurrence of argument-taking lexemes and various types of phrases (possible arguments), and the statistical inference stage at which reliable valence hypotheses are selected. Both stages will be discussed in detail. The course will conclude with the presentation of various evaluation methods and various uses of automatically extracted valence information. The closing bracket of the course will be the discussion of the extent to which automatic valence acquisition can help in distinguishing arguments from adjuncts.
Language and Computation advanced course:
Distributional Semantic Models - Theory and Empirical Results.
Teachers
- Stefan Evert ()
- Alessandro Lenci ()
Distributional semantic models (DSMs) are based on the
assumption that the meaning of a word can (at least to a certain
extent) be inferred from its usage, i.e. its distribution in
text. Therefore, these models dynamically build semantic
representations "in the form of multi-dimensional vector
spaces" through a statistical analysis of the contexts in
which words occur.
With their distributed vector-space representations, DSMs
challenge traditional symbolic accounts of conceptual and
semantic structures. However, their true ability to address key
issues of lexical meaning is still poorly understood, and will
have to be carefully evaluated in linguistic and cognitive
research.
This course aims to equip participants with the necessary
background knowledge for carrying out such research. In addition
to the mathematical foundations of DSMs and their application to
semantic analysis, we will put particular emphasis on relating
the computational models to fundamental issues of semantic
theory.
Language and Logic workshop:
Formal approaches to sign languages.
Organizers
- Carlo Cecchetto ()
- Carlo Geraci ()
Course material: 02 Brentari Eccarius ESSLLI 2009.pdf 03 Schlenker ESSLLI 2009.pdf 04 Koulidobrova ESSLLI 2009.pdf 05 Davidson Caponigro Mayberry ESSLLI 2009.pdf 06 Neidle et al ESSLLI 2009.pdf 08 MathurRathmann ESSLLI 2009.pdf 09 Bernaht ESSLLI 2009.pdf 10 Aboh Pfau 2009 ESSLLI.pdf 11 Donati Branchini ESSLLI 2009.pdf
Abstract:The recognition that sign languages are natural languages in their own right, and not collections of gestures or impoverished codes lacking an autonomous grammar, begins with Stokoe (1960). With Stokoe's work, the methods linguists use to describe and investigate spoken languages are applied to sign languages as well. In recent years, linguistic work on sign languages has also developed in formal frameworks, in the areas of phonology, syntax, and semantics. One goal of the workshop is to bring together researchers from different areas of formal linguistics who are investigating sign languages. The grammars of sign languages are as highly complex as the grammars of spoken languages and share with them many universal features, despite the difference in modality between spoken languages (which use the auditory channel) and sign languages (which use the visual channel). Yet, sign languages also differ from spoken languages in radical ways: morphological information in sign languages is often conveyed simultaneously by different articulators rather than linearly; moreover, certain aspects of their phonological, syntactic and semantic structures are not commonly found in spoken languages. These differences raise an interesting challenge for existing formal linguistic frameworks, which are designed to account for the grammars of spoken languages. By bringing together formal linguists working on sign languages, the workshop should contribute to meet this challenge.
Logic and Computation introductory course:
Logics and Agent Programming Languages
Teachers
- Natasha Alechina ()
- Brian Logan ()
Keywords: agent programming languages, logical models of agency, verification
Abstract:
To create a language for programming entities capable of intelligent behaviour (`agents'), researchers and developers must address deep questions such as: what are the basic constituent parts of an intelligent agent; how should the agent `think' (e.g., which deliberation strategy should it employ -- should it plan a precise sequence of actions in advance or should it adopt an abstract plans with gaps `to be filled-in later'); what relationship should there be between the agent's beliefs and its goals, etc. In seeking to address these questions, researchers have drawn heavily on formal models of agents and on agent logics, including epistemic logics, logics of action, dynamic logic, coalition logics etc. For example, the development of agent programming languages such as AgentSpeak were heavily influenced by the BDI (Beliefs, Desires and Intentions) logics developed to understand what an agent's behaviour should be. These interactions have resulted in an extremely fruitful cross fertilisation between work in logic and computation, and the application of logical techniques to address key practical issues such as the verification of agent programs (i.e., will an agent program meet the specification set out by its developers).
This course will address key topics in logics of agent programs including: the Belief Desire Intention model; overview of agent programming languages based on the BDI model; relationship between the operational semantics of BDI-based agent programming languages and logics for reasoning about agents' beliefs and intentions; verification of agent programs using model checking and theorem proving. The course assumes some exposure to modal logic, but no prior knowledge of agent programming languages is required.