Busemann, S. Best-first surface realization. In: Scott, D. (ed.) Proceedings of the 8th International Workshop on Natural Language Generation, University of Brighton, Herstmonceux Castle, UK, 1996. pp101110.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... related to focus of attention [8] In order to be coherent, a speaker takes into account the evolving attention focus of the hearer and presents his information from that perspective [9,10,11] Depending on the language, this process may involve, besides morpho syntactic, also prosodic features [12]. Generally, the linguistic context what was said before is assumed to determine the current attention focus. However, when generating verbal interaction on wearable devices, the physical context of the hearer obviously plays a role as well. Starting from these observations, we investigated and ....

Busemann, S. Best-first surface realization. In: Scott, D. (ed.) Proceedings of the 8th International Workshop on Natural Language Generation, University of Brighton, Herstmonceux Castle, UK, 1996. pp101110.

....1993] introduce pointers to KB individuals among fixed strings and insert canned text as the value of a frame representing the meaning of a sentence. Cancedda et al. 1997] insert templates as leaf nodes of a textual plan produced by classical NLG techniques. Hybrid techniques are also used by [Busemann 1996] and [Busemann and Horacek 1998] All these attempts to integrate templates within NLG architectures can be characterized as using Mixed Representations, which are absent from both pure NLG and static templates. Our approach to text generation aims at mixing representation levels in a systematic ....

Busemann S. Best-first surface realization. In INLGW-8, pages 101-- 110, Sussex, 1996.

....canned texts, and grammars are just generalized templates. Indeed, the possibility of combining these different modes of generation has recently been highlighted as one of the keys to efficient use of natural language generation techniques in practical applications (van Noord Neumann, 1996; Busemann, 1996). 6.3 Processing Let us now indicate how the rules are meant to be used by the generator module. Traditionally, the process of generation is divided into two steps: generation of message structure from database records (what to say) and generation of sentences from message structures (how to ....

Busemann, Stephan. 1996. Best-first surface realization. Computation and Language E-Print Archive cmp-lg/9605010. URL http://xxx.lanl.gov/cmp-lg/.

....of the context) and large linguistic resources (grammars and lexica) are necessary, which leads to a high initial overhead. To increase the flexibility of the generated texts, methods from linguistically motivated generation should be combined with the template based approach (see [38] and [4]) There are di#erent representation formalisms for describing the text structure, like static schema ( 17] or the functionaloriented rhetorical structure theory ( 16] Some text planning techniques (see for example [19] create an explicit representation of the text structure which cover the ....

S. Busemann. Best-First Surface Realization. In D. Scott, editor, Proc. of the 8th International Workshop on Natural Language Generation, Herstmonceux, Sussex, Great Britain, June, 13--15 1996.

....compensates for this limitation and we managed to reuse at least partly, some specific structures, e.g. for manipulating lists, or rendering dates, from an earlier application. Our NLG component is developed using the TG 2 tool 2 , which allows for template based NLG as well as deep generation [6]. We currently use templates only, but as we extend our system to more scenes, we are looking for generalities that are worth implementing small subgrammars, similar to the functions in the syntactic templates of [7] COMRIS text templates are encoded 2 TG 2 is used in the COMRIS project under a ....

Busemann, S.: Best-First Surface Realization. In: Scott, D. (ed) Proceedings of the 8th Intl. workshop on Natural Language Generation, Herstmonceux Castle, University of Brigthon, UK, (1996) 101-110

....NLG process history, physical context and user profile (Geldof, 1999a) Real time information about these different perspectives annotates the input structure of the template based NLG component. We use TG 2, a rule based engine that covers the continuum between templates and syntactic generation (Busemann, 1996). Making abstraction from planning and multi sententional discourse phenomena allows us to focus on the subject of our research: context sensitivity and surface form. In this paper, we want to uncover how context affects the structure of utterances (vs lexical choice) Section 2 presents the ....

S. Busemann. 1996. Best-first surface realization.

....of the context) and large linguistic resources (grammars and lexica) are necessary, which leads to a high initial overhead. To increase the exibility of the generated texts, methods from linguistically motivated generation should be combined with the template based approach (see [38] and [4]) There are di erent representation formalisms for describing the text structure, like static schema ( 17] or the functionaloriented rhetorical structure theory ( 16] Some text planning techniques (see for example [19] create an explicit representation of the text structure which cover the ....

S. Busemann. Best-First Surface Realization. In D. Scott, editor, Proc. of the 8th International Workshop on Natural Language Generation, Herstmonceux, Sussex, Great Britain, June, 13-15 1996.

....power, starting with inflexible canned methods and ending with maximally flexible feature combination methods. In the context of the MIETTA system the template generator JTG 2 from CELI DFKI will be adopted by the consortium. JTG 2 is a porting into Java TM of TG 2 (developed in LISP at DFKI [7]) a surface generator that can be easily hooked up to application systems or to language planning systems. In domains where no extensive utterance planning is needed, TG 2 can be parametrised according to preferred properties of utterances (style, wording, and contents) Parameters guide the ....

Busemann S., Best-first surface realization. In Donia Scott (Ed.), Eighth International Language Generation Workshop. Proceedings. 1996.

....linguistic knowledge. Productions are applied through the familiar three step processing cycle: i) identify the applicable rules, ii) select a rule on the basis of some conflict resolution mechanism, and (iii) apply that rule. Productions axe used to encode grammar rules in the lemguage TGL [Busemann, 1996]. A rule is applicable if its preconditions are met. The TGL rule in Figure 2 is applicable to input material as shown in Figure 1, because the COOPslot matches, and there is information about the THRESHOLD VALUE available (otherWise a different sentence pattern, and hence a different rule, ....

....simple template filling techniques. A TGL rule is successfully applied if all actions are carried out. The rule returns the concatenation of the substrings produced by the template actions. f an action fails, backtracking can be invoked fiexibly and eciently using memoization techniques (see [Busemann, 1996]) 4 Costs and Benefits As Reiter and Mellish note, the use of shallow techniques needs to be justified through a cost benefit analysis [Reiter and Mellish, 1993] We specify the range of possible applications our approach is useful for, exemplified by the report generator developed for the ....

[Article contains additional citation context not shown here]

Stephan Busemann. Best-first surface realization. In Donia Scott, editor, Eighth International Natural Language Generation Workshop. Proceedings, pages 101-110, Herstmonceux, Univ. of Brighton, England, 1996. Also available at the Computation and Language Archive at http://xxx. lanl. gov/abs/cmp-lg/9605010.

....linguistic oriented input to the generator. The same holds for failure messages, such as (02) and for specifications of free time slots, as in (07) where simple rules of aggregation take care not to repeat the full date specification for each clock time mentioned. The production system TG 2 [Bus96] proved to be sufficiently flexible to accomplish this task by its ability to generate preferred formulations first. For instance, Cosma clients can parameterize TG 2 so as to refer to their owner by a first person pronoun or by a full name, or to use formal or informal form of addressing the ....

Stephan Busemann. Best-first surface realization. In Donia Scott, editor, Eighth International Natural Language Generation Workshop. Proceedings, Herstmonceux, Univ. of Brighton, England, 1996. Also available as Research Report RR-96-05, Deutsches Forschungszentrum fur Kunstliche Intelligenz, Saarbrucken, Germany.

.... Language Dialogue Service for Appointment Scheduling Agents Stephan Busemann, Thierry Declerck, Abdel Kader Diagne, Luca Dini, Judith Klein, Sven Schmeier DFKI GmbH Stuhlsatzenhausweg 3, 66123 Saarbrucken, Germany busemann dfki.uni sb.de Abstract Appointment scheduling is a problem faced daily by many individuals and organizations. Cooperating agent systems have been ....

Stephan Busemann. 1996. Best-first surface realization.

....is characterized by a simple and small sublanguage. The system is called with a bundle of user requests entered through a hyper link navigator. For text planning, a schema based component produces domain specific semantic content representations that are fed to the TG 2 production system [ Busemann, 1996 ] for linguistic realization. The semantics interface between the two components is tailored to the task and domain at hand. It is independent from the particular language chosen. It is argued that these design decisions have important practical benefits over more general, linguistic approaches. ....

....The generation system must fulfill two tasks: Text structuring: A schema based component generates the text structure on the basis of the user s request. It combines fixed text blocks with dynamic text in an as languageneutral way as possible. Surface realization: The production system TG 2 [ Busemann, 1996 ] is reused for languagespecific processing. It will operate with grammars partly tuned towards the domain and task requirements. In addition, in order to be cooperative and helpful, the system must exhibit robustness in all its parts. Any of the interfaces in the generator may provide unexpected ....

[Article contains additional citation context not shown here]

Stephan Busemann. Best-first surface realization. In Donia Scott, editor, Eighth International Natural Language Generation Workshop. Proceedings, Herstmonceux, Univ. of Brighton, England, 1996. Also available as Research Report RR-96-05, Deutsches Forschungszentrum fur Kunstliche Intelligenz, Saarbrucken, Germany.

No context found.

Stephan Busemann. 1996. Best-first surface realization.

No context found.

Stephan Busemann. Best-first surface realization. In Eighth International Natural Language Generation Workshop, pages 101--110, Sussex, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC