Call for Papers: ECAI 2010 Workshop
Language, Pragmatics and Explanation, LPE’10
Lisbon, August 2010
Outline and aims
Explanation analysis and production is a wide topic that has been addressed with specific objectives in disciplines such as didactics and psychology, and in domains as diverse as law and legal reasoning, politics and news editorials production, human resource management, etc. Some forms of empirical analysis have emerged from these areas, often under the form of schemas that characterize certain forms of explanations and their expected role.
Some useful work has been carried out so far on formal aspects: mental and conceptual models, semantics and pragmatics of explanation, but less work on AI and related language technology (lexicon, grammar, and semantics) aspects. Besides the need of formal developments, computer-based explanation production is now becoming a crucial component of a number of application areas, as, for example, cooperative question-answering, involving language processing and reasoning, where most responses provided by a system need to be paired with explanations on how the response has been found and on how it responds to the question. Lets us also cite computer-based didactic systems where the transmission of a certain knowledge or know-how needs to be paired with quite sophisticated forms of explanations. In most contexts, these explanations must be generated dynamically: they cannot be defined a priori via a few simple forms. On a more abstract register, explanation analysis is crucial in a number of situations, such as opinion analysis from news editorials or political speeches, where the relevance and quality of the explanations will heavily influence the opinions of listeners. Finally, procedures (maintenance, social life, DIY, etc.) are a rich source for explanation analysis and production.
Recent foundational, methodological and technological developments in knowledge representation, in pragmatics (user models and profiling, intentions modelling, etc.), in reasoning (e.g. argumentation, decision theory, etc.), in advanced language processing, including semantic and discourse analysis, and recent progress in Human Language Technology, make it possible to foresee the elaboration of much more accurate, cooperative and robust systems dedicated to explanation analysis and production. These can be essentially language-based (via online texts or web pages, operating either on open or closed domains) or they can incorporate multimedia aspects. The user interface aspects (input, output (e.g. SMS or advanced interfaces), on line help, dialogue, etc.) are also crucial in the definition of such systems. This leads us to also welcome contributions that involve multi-media aspects.
The workshop addresses both empirical and formal models of explanation. It includes, but is not limited to, the following topics:
Linguistic and conceptual aspects
- Rhetorical structures that support or contribute to explanation production and structure (elaborations, reformulations, examples, warnings, hints, various forms of arguments, etc.) and their structure and semantic and pragmatic role,
- Illocutionary force identification, persuasion effects,
- The language (lexical, grammatical) of explanation, language connectors, metaphors in explanation,
- Analysis of surface forms of explanation: textual, iconic, typography, graphics, images, multimodality,
- Characterization of explanation patterns (e.g. sequences of rhetorical structures) and their communicative role.,
- Specific problems, such as e.g. paraphrase and explanation,
- Cognitive aspects: perception of explanations by readers, implicature,
- Philosophical aspects and Ethics of explanation.
Annotation techniques and guidelines
- Guidelines for annotating explanation segments: what to annotate, and how,
- Annotator training aspects,
- Corpus development and validation: parameters to consider, various forms of corpora, tools for corpus collection, analysis and validation technologies.
Language processing aspects :
- Identification of explanation structures: by what means, lexical and grammar resources required, processing strategies,
- Resource and portability issues: the lexicon of explanation, domain related grammatical structures,
- Semantic representation, declarative and ‘operational’ semantics of connectors and rhetorical operators
- Specific linguistic aspects, such as: explanation and paraphrasing.
Artificial intelligence aspects
- The logic of explanation,
- Models for explanations,
- Explanation-based systems, explanation in expert systems,
- Implicature, textual entailment, logic of assumptions and presuppositions,
- Related knowledge representation aspects, analysis of the required domain knowledge.
Explanation production, natural language generation issues
- Planning issues, e.g. w.r.t. explanation goals,
- Implementation of pragmatic aspects such as cooperativity,
- Lexical choice, sentence planning, taking into account of the user profile,
- Architecture of systems that support explanation generation
Applicative domains, Man-Machine Communication, multimedia
- Analysis of explanation structures per domain, analysis of their role, their perception by users, specificities of certain domains and audiences
- Experiences from applications, e.g. as in: didactics (of science, of second language acquisition, etc., ); procedures and procedural knowledge; law; public debates, news and editorials; business; philosophy, etc.
- Explanation and its underlying effects: persuasion, emotions,
- Granularity of explanations, focusing, zooming in/out, user interactions and explanation management
- Multimedia aspects of explanation: e.g. pairing videos with textual explanations.
Evaluation techniques and methods
- Identification of features to evaluate: technology aspects, user interaction and interpretation aspects, protocol definition and relevance measures,
- Engineering aspects: portability, re-usability, robustness, development costs and resources, interoperability, risk analysis,
- Explanation and economical models.
The workshop will be organized around a few major questions of interest to a number of AI, NLP, linguistics, HLT, psychology and pragmatics people. Practitioners of explanation in various areas such as didactics, opinion analysis, experimental psychology, ethics, are encouraged to participate. Application developers are also most welcome.
The goal of this workshop is to enhance cross-discipline interactions. Contributors must be opened to interactions with the different workshop areas. The programme committee will care to have a balanced number of participants from the different areas concerned.
Paper will go under peer review, as for most conferences.
To encourage an atmosphere appropriate
for a workshop, we plan to:
- have a 15mn discussion at the end of each session,
- have a panel and an invited speaker, possibly plan a tentative roadmap definition,
- plan demos on portable machines.
We welcome short papers (max 3 pages), describing projects or ongoing research and long papers (max. 6 pages), that relate more established results. Papers describing applications are encouraged, besides theoretical contributions. Papers must be sent in .pdf format. The format to use for papers and abstracts is the same as for ECAI. The title page (no separate title page is needed) should include the following information:
Papers must be sent to both co-organizers at the email addresses given below.
May 7th 2010: paper submission
June 7th: acceptance/rejection notification
June 25th: reception of final paper
All accepted papers (long and short) will be published in the workshop proceedings.
Leila Amgoud, Patrick Saint-Dizier CNRS
IRIT 118 route de Narbonne 31062 Toulouse cedex France.
Phone : +33 5 61 55 62 44/
Note: all workshop participants are expected to register to the workshop and to the main conference.