Elhadad, Michael. 1992. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. Ph.D. thesis, Columbia University, New York, NY. (The SURGE grammar for FUF is available via anonymous FTP at: ftp://www.cs.columbia.edu/fuf).  Home/Search   Document Details and Download   Summary   Related Articles   Check

....cfi he objects, and on generating references to the presentation itself. Previous work 0n reference in sentence generation, e.g. lAppelt1985, Dale1992, Dale and Reiter1995, Heeman and Hirst1995, Horacek1997, Stone and Doran1997] has not addressed the referential attributive distinction. Elhadad1992] describes a representation scheme for specifying complex noun phrases, in which a set can be described either by its extension or intension. However, this distinction.is independent of the referential attributive distinction, since the same noun phrase can be used with either intention. 7 ....

M. Elhadad. 1992. Using Argumentation to control Lexical Choice: A Functional Unification 'Implementation. Ph.D. thesis, Columbia University.

....intended structure. The other central component, the discourse planner, is responsible for selecting the communicative techniques that will be used to convey the unfolding action to the user. We briefly describe these two components here. Additional components, such as a text generation system [3] (used to create custom dialog and narration) a user model and an HTTP server are also included in the MC; space limitations preclude discussion of the design of these elements. Both the narrative and discourse planner components use the Longbow planning system, a hierarchical partial order ....

Elhadad, M. (1993). Using argumentation to control lexical choice: A functional unification based approach. Ph.D. Thesis, Columbia University.

....to each object will be generated as well and used as labels on the illustrated object in the accompanying graphics. Thus, for example, speech just says Two IVs while the textual labels spell out that they are peripheral. MAGIC uses Columbia s FUF SURGE language generation tools for this task [8, 9, 10]. RELATED WORK A key feature of MAGIC is that both content and form of the presentation are dynamically generated at runtime, thus allowing output to be customized for the current user and situation. In contrast, many interfaces for healthcare applications rely on the more traditional approach ....

M. Elhadad. Using argumentation to control lexical choice: a unification-based implementation. PhD thesis, Dept. of Computer Science, Columbia University, 1993.

....words are selected to realize the values of the fields. The resulting case frame is passed through syntactic generation, where a full syntactic tree of the sentence is created, grammatical constraints are enforced, and morphological agreement is carried out. We use Columbia s FUF SURGE package [5, 16], a robust grammar of English (SURGE) along with a unification interpreter (FUF) and some text manipulation tools written in Perl as the basis for both word choice and syntactic generation. 6 Conclusions and Status In the CDNS project, we have developed tools and a system architecture with which ....

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Aho et al.

....(automatically retrieved) collocations in text generation. 10 The Xtract system mentioned above (Section 15.3) used statistical techniques to extract collocations from free text. The output of Xtract was then fed to a separate program, Cook [39] which used a functional unification paradigm FUF [26, 16]) to represent collocational knowledge, and more specifically, constraints on text generation imposed by the collocations and their interaction with constraints caused by other components of the text generation system. Cook could be used to represent both compound collocations (such as the Dow ....

Michael Elhadad. Using argumentation to control lexical choice: a unification-based implementation. PhD thesis, Computer Science Department, Columbia University, 1993.

....can be used in places. Sentence planning helps create a text that 10 will be easier for readers to comprehend. In general, there are three subtasks involved in sentence planning [29] Lexicalization chooses the particular words or phrases that are required to communicate the speci ed content [3]. The selection of words is based on pragmatics, style, and user modeling [27] Aggregation decides how parts of text should be combined into sentence sized chunks. It partially decreases redundancy in a text. Referring expression generation determines the use of a pronoun or a phrase that is ....

....of a grammar. 16 ( cat clause) tense present) subject ( lex John ) number singular) np type proper) verb read ) object ( lex book ) definite yes) number singular) Figure 2.5: A Feature Structure of The Sentence John reads the book. PENMAN NIGEL [19] and FUF SURGE [3, 4] are examples of feature based text realization systems that are widely used in the NLG research community. They are di erent in their underlying implementation of grammar and the feature collection process. PENMAN NIGEL uses a Systemic Grammar and proceeds by traversing feature selection ....

[Article contains additional citation context not shown here]

Michael Elhadad. Using argumentation to control lexical choice: A functional unication-based approach. PhD thesis, Computer Science Department, Columbia University, 1992.

....to, that routinely produced by human authors. This architecture has been implemented in STORYBOOK, an end to end narrative prose generation system that utilizes narrative planning, sentence planning, a discourse history, lexical choice, revision, a full scale lexicon, and the well known FUF SURGE [Elhadad, 1992] surface realizer to produce multi page stories in the Little Red Riding Hood fairy tale domain. ONCE UPON A TIME GEORGE ANT LIVED NEAR A PATCH OF GROUND. THERE WAS A NEST IN AN ASH TREE. WILMA BIRD LIVED IN THE NEST. THERE WAS SOME WATER IN A RIVER. WILMA KNEW THAT THE WATER WAS IN THE ....

....Ordering: Many event verbs also impose different theta frames even though they describe similar actions. Little Red Riding Hood might give her grandmother the cookies or grandmother might receive the cookies from Little Red Riding Hood. Although this does not compare to more sophisticated methods [Elhadad, 1992; Stede, 1996 ] and is by no means suggested as a solution to the problem of lexical choice, it is sufficient to satisfy our goal of preventing repetitive prose. The result of segmentation, discourse history analysis, and lexical choice is thus a modified narrative stream. However, in classic ....

[Article contains additional citation context not shown here]

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Columbia University, 1992.

....of generation depends on the complexity of the template that the application selects, not on the size of the grammar. Short, simple utterances are always realized faster than longer ones. In many other approaches, speed is a function of the grammar size, because it is searched during realization [5, 6, 9, 10, 11]. Robustness In YAG, the quality of text is dependent on the quality of templates and also the completeness of input. However, the realization of a template cannot fail. Even if there are inconsistencies in its input (such as subject verb disagreement) the generator will produce an ....

....it is possible to de ne a grammar using TG 2 s rules, its current distribution does not attempt to do so. Most generated texts are canned with little grammatical information. In contrast, YAG provides a general purpose template based grammar that can be reused across domains. 2 FUF SURGE [5, 6] uses a functional uni cation grammar to realize text. The realization process uni es a given input with the grammar, and generates text from the uni cation result. SURGE is an extensive grammar of English that is provided with the system. Other systems, such as Penman Nigel [9] do not use ....

Michael Elhadad. Using argumentation to control lexical choice: A functional unicationbased approach. PhD thesis, Computer Science Department, Columbia University, 1992.

.... represented as a series of I1 classes, and classification is used to apply the grammar to the SPL produced by the sentence planner [ Reiter and Mellish, 1992; Mellish, 1991 ] The idas grammar is small when compared to, for example, ISI s nigel [ Mann, 1983 ] grammar or Elhadad s surge grammar [ Elhadad, 1992 ] but it is adequate for idas s needs (remember that writing standards for technical documentation generally prohibit complex syntactic structures in any case) Morphology. Morphological processing in idas is again done with classification; some of the specific rules are taken from [ Ritchie ....

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Columbia University, 1992.

....in a way that preserves fluency Simply stringing the phrases together can produce ungrammatical results because phrases are placed in new syntactic contexts. We have developed an algorithm that maps the predicate argument structure from input document phrases to arguments expected by FUF SURGE (Elhadad 1993; Robin 1994) a robust language generation system. This has required developing new techniques to identify constraints on realization choice (e.g. on the order of circumstantial roles such as time, location, instrument, etc. using surface features in place of the semantic or pragmatic ones ....

....should be combined into a single, more complex sentence, looking again at constraints from the input document as well as common references between phrases. Finally, the constituent structure produced by these two stages is mapped to the functional representation required as input by FUF SURGE (Elhadad 1993; Robin 1994) Document Analysis The first stage of our system identifies similar text units; that is, those that contain important common information. Our definition of similarity is di#erent than the one adopted in most text matching tasks (such as information retrieval) because of two ....

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Department of Computer Science, Columbia University, New York, 1993.

....similar phrases in novel contexts Simply stringing the phrases together can produce ungrammatical results because phrases are placed in new syntactic contexts. We have developed an algorithm that maps the predicate argument structure of input document phrases to arguments expected by FUF SURGE [Elhadad 1993; Robin 1994] a robust language generation system. This has required developing new techniques for identifying constraints on realization choice (e.g. on the order of circumstantial roles such as time, location, instrument, etc. using surface features in place of the semantic or pragmatic ....

....should be combined into a single, more complex sentence, looking again at constraints from the input document as well as common references between phrases. Finally, the constituent structure produced by these two stages is mapped to the functional representation required as input by FUF SURGE [Elhadad 1993; Robin 1994] Document Analysis Our definition of similarity is different than the one adopted in most text matching tasks (such as information retrieval) because of two factors: first, the size of the unit of text affects what is similar; documents have a lot of information, so even a modest ....

[Article contains additional citation context not shown here]

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Department of Computer Science, Columbia University, New York, 1993.

....determines the phrases and surface syntactic form of the summary. The linguistic component consists of a lexical chooser, which determines the high level sentence structure of each sentence and the words which realize each semantic role, and the fuf surge sentence generator [Elhadad, 1991, Elhadad, 1993] As mentioned earlier, input to summons is a set of templates, where each 18 template represents the information extracted from one or more articles by a message understanding system. We should note that we based our implementation both on output templates from actual muc systems and on ....

....used to used to produce templates from all related articles. Articles are then grouped according to topic [Radev et al. 1999] into clusters. Each of the clusters becomes the input of the text generation module of summons (Figure 2. 4) That gure is partially adapted from Michael Elhadad s thesis [Elhadad, 1993]. In addition to the conceptual operators described elsewhere, what makes 20 News source 1 AFP News source 2 Reuters News source n Times . Clusters . Templates T 1 T 2 T 3 T 4 T 5 T n 1 T n C C 2 C 3 C 4 C 5 T 1 T 2 T 3 T 4 T 5 T n 1 T n Figure 2.3: Online processing: ....

[Article contains additional citation context not shown here]

Michael Elhadad. Using argumentation to control lexical choice: a unication-based implementation. PhD thesis, Computer Science Department, Columbia University, 1993. 166

....Together, they have the potential of generating better speech than Text to Speech (TTS) systems. In this paper, we present a series of experiments that use machine learning to identify correlation between intonation and features produced by a robust language generation tool, the FUF SURGE system (Elhadad, 1993; Robin, 1994) The ultimate goal of this study is to provide a spoken language generation tool based on FUF SURGE, extended with an intonation generation component to facilitate the development of new CTS applications. 2 Related Theories Two elements form the theoretical background of this ....

M. Elhadad. 1993. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. Ph.D. thesis, Columbia University.

....as opposed to written language. 2. RELATED WORK Our research builds on techniques developed for language generation [9] incorporating components that handle the tasks of content planning, lexical selection, and syntactic sentence generation. Our speech generator uses the Fuf Surge package [6], a sentence generation system containing a robust English grammar. Unlike previous work in language generation, our work focuses on the development of techniques for producing spoken language. We used Lucent Bell Laboratories Text To Speech (TTS) system for speech synthesis. For a general ....

M. Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Columbia University, 1993.

....many of the parameters our CTS system address are those needed for aiding coordination between media. Currently, there are three components in the CTS system: an NLG component, a set of CTS algorithms and a TTS. The NLG tools and TTS are application independent. We use the FUF SURGE package (Elhadad, 1993) for generation. The speech synthesis system is AT T Bell Labs TTS system. The concept to speech algorithms, however, are not system independent. The input of these algorithms are in FUF SURGE representation and the output is designed specificly for AT T TTS input. In this section we describe our ....

M. Elhadad. 1993. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. Ph.D. thesis, Columbia University.

....once the tool is designated as UNI NO QUEUE. 5.3. MULTI QUEUE: FUF FUF is a sophisticated unification based tool running on top of Lisp and is used, among other things, in natural language processing research for the generation of sentences from corresponding syntactic data structures [23]. It defines hierarchical procedures that apply in sequence one or more separate layers of unification rules to its input structures as well as to the new structures produced by each step of the procedure in order to obtain as output all the valid surface forms, under the constraints posed ....

Michael Elhadad. Using argumentation to control lexical choice: a unification-based implementation. PhD thesis, Columbia University, Department of Computer Science, 1993.

....in most knowledge bases and the representational language used to encode linguistic knowledge in sophisticated surface realizers. Knowledge bases typically employ a logic formalism, such as a subset of FOPC provided by a frame based system. In contrast, the state of the art surface realizer, Fuf [11, 12], employs a unification based formalism to encode input, grammatical knowledge, and lexical knowledge. In principle, these languages could be very similar; in practice, they differ radically. As a result, bridging the gap between domain knowledge and linguistic knowledge has become a critical ....

....of legibility. 4 Functional descriptions such as the one in the figure are produced by a functional realizer and passed to a unification based surface generator for realization as text. They have proven to be very effective for surface realization. For example, the Fuf surface realization system [11, 12] employs functional descriptions to represent both its input and its grammatical and morphological knowledge. Fuf, which was developed by Elhadad and his colleagues at Columbia, is the state of the art surface realization system. It is accompanied by an extensive, portable English grammar, Surge, ....

[Article contains additional citation context not shown here]

Michael Elhadad. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. PhD thesis, Columbia University, 1992.

....(that decides how to say it) Recent research has led to the development of reusable surface realization systems which encapsulate knowledge of syntax for specific languages. Most of this research has been conducted for English, and large generation grammars for English are now available [4], 9] Multi lingual generation now appears as the next frontier in generation research: it is often easier to develop applications to generate text in different languages than it is to translate text. When developing a multi lingual application, the engineering strategy is to develop a single ....

....one of all, both, half, multipliers and fractions, der is a deictic determiner, deictic is an adjective from a restricted class that expresses the anaphoric status of the thing referred to (e.g. above, same, different) and quantafter expresses the amount or quantity of the thing referred to. In [4], we gathered a minimal set of 24 features to define the determiner sequence of the NP in English. Complex systems in SURGE implement the cooccurence restrictions among these features, definiteness and partitive constructs. 92 Definite yes no The a book Distance far near none This That The ....

Michael Elhadad. Using argumentation to control lexical choice: a unification-based imple- mentation. PhD thesis, Computer Science Department, Columbia University, 1992.

....of the input specification language to a syntactic realization component through a contrastive analysis of .the requirements of English and Hebrew. By design, we are attempting to keep the input to HUGG as similar as possible to the one we defined in the SURGE syntactic realization for English [7]. A detailed 168 analysis of syntactic constructs specific to Hebrew becomes, therefore, critical to evaluate to which extent the input specification language can abstract away from knowledge of the syntax. We investigate in this paper one such construct: the Hebrewnoun compounding form known as ....

Michael Elhadad. Using argumentation to control lexical choice: a unification-based imple- mentation. PhD thesis, Computer Science Department, Columbia University, 1992.

No context found.

Elhadad, Michael. 1992. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. Ph.D. thesis, Columbia University, New York, NY. (The SURGE grammar for FUF is available via anonymous FTP at: ftp://www.cs.columbia.edu/fuf).

No context found.

M. Elhadad. Using Argumentation to Control Lexical Choice: AFunctional Uni#cation Implementation. PhD thesis, Columbia University, 1993.

No context found.

Elhadad, M. 1992. Using Argumentation to Control Lexical Choice: A Functional Unification-Based Approach. Ph.D. diss, Columbia University.

No context found.

Michael Elhadad. 1993. Using Argumentation to Control Lexical Choice: A Functional Unification Implementation. Ph.D. thesis, Department of Computer Science, Columbia University, New York.

No context found.

Michael Elhadad. Using argumentation to control lexical choice: A functional unificationbased approach. PhD thesis, Computer Science Department, Columbia University, 1992.

No context found.

Michael Elhadad. 1992. Using Argumentation to Control Lexical Choice: A Functional Unification-Based Approach. Ph.D. thesis, Department of Computer Science, Columbia University.

First 50 documents  Next 50

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