Structural priming in language production is a tendency to recreate a recently uttered syntactic structure in different words. This tendency can be seen independent of specific lexical items, thematic roles, or word sequences. Two alternative proposals about the mechanism behind structural priming include (a) short-term activation from a memory representation of a priming structure and (b) longer term adaptation within the cognitive mechanisms for creating sentences, as a form of procedural learning. Two experiments evaluated these hypotheses, focusing on the persistence of structural priming. Both experiments yielded priming that endured beyond adjacent sentences, persisting over 2 intervening sentences in Experiment 1 and over 10 in Experiment 2. Although memory may have short-term consequences for some components of this kind of priming, the persisting effects are more compatible with a learning account than a transient memory account. Speakers repeat themselves. Sometimes their repetitions are intentional, made for emphasis or other stylistic and social purposes (Giles & Powesland, 1975; Tannen, 1987), and sometimes they are accidental. They may involve almost

The acceptance of natural language generation systems strongly depends on their capability to facilitate the exchange of information with human users. Current generation systems consider the influence of situational factors on the content and the form of the resulting utterances. However, the need to time their processing flexibly is usually neglected although temporal factors play a central part when directly addressing a human communication partner. A short response time of a system is crucial for its effective use. Furthermore, some applications --- e.g., the simultaneous description of ongoing events --- even necessitate the interleaving of input consumption and output production, i.e. the use of an incremental processing mode. We claim that incremental processing is a central design principle for developing flexible and efficient generators for speech output. We discuss the advantages of parallel processing for incremental generation and several aspects of control of the generator...

Exploiting an incremental and parallel processing scheme is useful to improve the performance of natural language generation systems. TAG--GEN is a TAG--based syntactic generator that realizes both principles. It is shown how the demands of incremental and parallel generation influence the definition, the design, and the processing of syntactic rules on the basis of Tree Adjoining Grammars. Keywords Natural Language Generation, Incremental Processing, Tree Adjoining Grammar 1 Introduction The growing importance of natural language for human--computer interfaces motivates researchers in the field of artificial intelligence to search for methods that improve the performance of those systems. We developed the flexible and efficient syntactic generator TAG-- GEN as part of the text--generation module of the multi--modal presentation system WIP 1 (cf. [Wahlster et al. 92], [Wahlster et al. 93]). The design of TAG--GEN has been influenced by a number of goals (see [Kilger & Finkler 93])...

Natural language reports generated by existing systems ignore the historical context of the facts and events they relate. In this paper, I argue that going beyond this limitation requires abandoning the pipelined architecture of existing report generators. I propose a new architecture in which a first draft of the report is organized around new information and then incrementally revised to opportunistically add related historical information. This type of information-adding revision allows to elaborate inside a clause or a nominal while taking into account surface structure constraints from any other portion of the report. In addition to providing the additional flexibility required to convey historical information, the proposed architecture constitutes an interesting testbed to investigate a wide range of open questions concerning content planning below the sentence level, generation with revision and generation architecture. 1 Introduction: generating reports in their historical co...

Abstract not found

This article reports 7 experiments investigating whether utterances are planned in a parallel or rightward incremental fashion during language production. The experiments examined the role of linear order, length, frequency, and repetition in producing Dutch verb-particle combinations. On each trial, participants produced 1 utterance out of a set of 3 as quickly as possible. The responses shared part of their form or not. For particle-initial infinitives, facilitation was obtained when the responses shared the particle but not when they shared the verb. For verb-initial imperatives, however, facilitation was obtained for the verbs but not for the particles. The facilitation increased with length, decreased with frequency, and was independent of repetition. A simple rightward incremental model accounts quantitatively for the results. Skilled behavior such as speech production requires advance planning of action components (e.g., Lashley, 1951). In the present paper, I examine what mode the speech production system uses in planning the morphophonological form of an utterance. In particular, I investigate whether the

To produce good utterances from non-trivial inputs a natural language generator should consider many words in parallel, which raises the question of how to handle syntax in a parallel generator. If a generator is incremental and centered on the task of word choice, then the role of syntax is merely to help evaluate the appropriateness of words. One way to do this is to represent syntactic knowledge as an inventory of “syntactic constructions” and to have many constructions active in parallel at run-time. If this is done then the syntactic form of utterances can be emergent, resulting from synergy among constructions, and there is no need to build up or manipulate representations of syntactic structure. This approach is implemented in FIG, an incremental generator based on spreading activation, in which syntactic knowledge is represented in the same network as world knowledge and lexical knowledge.

Summarization applications raise several challenging issues for language generation systems. To address them, I propose a new generation model where an initial draft conveying only the essential information is incrementally revised to include additional background information. The generator streak, implementing this model, relies on revision operations specifying the various ways a draft can be transformed in order to concisely accommodate a new piece of information. These operations are, for the most part, domain-independent. 1 Introduction In recent years, the volume of information available on-line has grown exponentially, and several large-scale efforts, such as the information superhighway and digital library initiatives in the US, are currently under way to further accelerate this growth. In order to put this abundance of electronic data to good use, and avoid information inundation, the development of automatic summarization facilities is critical. On-line information comes i...

Natural language generation must work with insufficient input. Underspecifications can be caused by shortcomings of the component provkling the input or by the preliminary state of incrementally given input. The paper Mms escape from such dead-end situ,tions by making kssumptions. We discuss global aspects of default lmndling. Two problem classes for deftmits in the incremental syntactic gonor;[tor VM-GI]N m'e presented to substantiate our discusskin.

In this paper I investigate the issue of providing historical background in computer-generated reports. I first observe that ignoring this issue is the most drastic limitation of existing report generation systems. I then present an empirical corpus analysis of basketball summaries aimed at discovering the specific means by which historical information is conveyed in human-generated reports. This analysis resulted in a set of data that forms the basis for the implementation of streak, a system generating basketball report leads with historical information. This data shows that building such a system cannot be based on the single-pass pipelined architecture of previous systems. Instead, I propose an entirely new architecture in which generation proceeds in two passes. The first pass builds a report draft containing only the basic facts. The second pass incrementally revises this draft to include additional facts providing the historical background. Independently of the issue of...