Knott, A.: 1996. A Data-Driven Methodology for Motivating a Set of Coherence Relations. PhD Thesis. University of Edinburgh, Edinburgh.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....directly or through bridging (as with too, otherwise, then, for example) 19 21] In addition, some connectives can convey more than what they assert, by means of implicatures. For example, negative polarity causal connectives like although, however, nevertheless can implicate a denied expectation [4, 5, 9 11]. We show that IS a ects both what is implicated by a connective (e.g. although) and what is available as an antecedent for an anaphoric connective (e.g. otherwise) Consider the di erences between (a) and (b) below. Small capitals indicate words carrying pitch accents, i.e. phonologically ....

Alistair Knott. A Data-driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, Department of Articial Intelligence, University of Edinburgh, 1996.

....defines a domain of structural locality only when such domains do not cross tree nodes. 3 A Case Study: However For the connective however, our first assumption is that however anchors the structural auxiliary tree, shown in Figure 1. 3(a) Regarding its semantic contribution, we follow (Knott, 1996) and 3 However, until we have accumulated ample empirical evidence, such conclusions are tentative and subject to revision. Also, it would be very interesting to investigate languages which allow crossed dependencies at the sentence level (e.g. Dutch) and examine whether in those languages ....

....inter relations between them. For example, in (11) not only does the clause medial position of however, flag then as a contrastive theme (in contrast with alternatives provided in the discourse or the speech situation) but the defeasible rule presupposed (or conventionally implicated) by however (Knott, 1996; Lagerwerf, 1998) involves that specific inertial property i.e. if someone smiles, they will continue to do so. However asserts that it fails to hold, and what happened then is the source of the failure. While we have not yet explored this with respect to LTAG and D LTAG, Bierner Webber, ....

Knott, Alistair (1996). A Data-Driven Methodology for Motivating a Set of Coherence Relations, (Ph.D. thesis). Edinburgh: University of Edinburgh.

....tree for the presentational relations appears in figure 1. When marking these relations, remember to follow the tree, and at the end confirm your choice by checking the relation definition and the examples. The definitions for these relations come from [4] The discourse cues come from [4] and [2]. Antithesis Conditions: 1. The situations of N and S are in contrast. 2. One of the di#erences between N and S means that N and S can t simultaneously be considered to be true, valid, good or preferable. 3. The speaker views the situation of N as true, valid, good or preferable. 4. Knowing S, ....

....decision tree for these relations starts in figure 2. When marking these relations, remember to follow the tree, and at the end confirm your choice by checking the relation definition and the examples. Most of the definitions for these relations come from [4] The discourse cues come from [4] and [2]. 15 Abstract:instance Conditions: 1. S presents an instance of, or example of, the concept described in N. Contrast with: set:member. Cues: for instance , for example Example 1 (N) The pongid family of the order primate is closely related to mankind. S) A particularly interesting ....

[Article contains additional citation context not shown here]

A. Knott. A Data-Driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, Department of Artificial Intelligence, University of Edinburgh, 1996.

....or parts of discourse. Standard examples are oppositive connectives (however, yet, etc. rephrasing connectives (in other words, in short, etc. or consequence connectives (so, therefore, etc. Quite a few studies have been devoted to them. See for example (Reichman 1985, Schiffrin 1987, Knott 1996) for English, Lang 1984, Grote and Lenke 1995) for English and German, Roulet et al. 1985, Jayez 1988a, Moeschler 1989) for French, Rossari 1994) for French and Italian, Ferrari 1995) for Italian. The background perspective of this paper is that of computer assisted language learning (CALL, ....

Knott, Alistair. 1996. A Data-Driven Methodology for Motivating a Set of Coherence Relations. Ph.d. thesis, University of Edinburgh.

.... of rules (section 3) second the type of update (section 4) 3 Admissible rules It has been repeatedly observed that DM can link entities of various semantic types, such as propositional contents (4 a) beliefs or judgments (4 b) and utterances or speech acts (4 c) see Sweetser 1990, Knott 1996, Sanders 1997 and Degand 1998 for recent presentations) 4) a. Mary was late, so she was cranky b. Mary was cranky. Then, she was probably late c. Where is the hammer, cause I ve to fix the shelf Although the intuitive fact is uncontroversial, it turns out, on closer inspection, that the ....

Knott, A. (1996). A Data--Driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, University of Edinburgh.

....certain proofs, which capture some aspects of the inferential properties exhibited by inferential PCs. We begin by explaining intuitively the difference between inferential PCs and the other two classes (section 2) Turning to inferential PCs in 1 See for example (Reichmann 1985; Schiffrin 1987; Knott 1996) for English, Lang 1984; Grote Lenke 1995) for English and German, Roulet et al. 1985; Jayez 1988; Moeschler 1989) for French, Rossari 1994) for French and Italian, Ferrari 1995) for Italian, Gil 1995; Fernandez 1996) for Romance languages. Note that the terminology is not fixed: in the ....

....contents of the sentences they connect. For instance, a contrast like (6) above would get a representation similar to (6 ) 8 6 See Konig (1986) who says There are many cases of overlap and neutralization so that a watertight system of classification and analysis does not seem to be possible. (Knott 1996) contains an interesting discussion of the classification problem and proposes an empirical method to address it. 7 Note that there is more to but than just the idea of a contrast. 8 We use x; y etc. as variables for non temporal entities, e and s for events and states, now for the time of ....

[Article contains additional citation context not shown here]

Knott A. (1996) A Data--Driven Methodology for Motivating a Set of Coherence Relations, Ph.D. thesis, University of Edinburgh.

....0 stand for the propositions expressed by and , stands for defeasibly implies in Asher and Morreau s common sense entailment [Asher and Morreau, 1991] and gen(X) stands for a generalization, i.e. an abstraction reachable from proposition X. A similar treatment appeared independently in [Knott, 1996] and [Knott and Mellish, 1996] Lagerwerf applied the standard presupposition tests (including embedding under negation or modal operators) and discourse context tests. The 5 latter rely on the notion of satisfaction of presuppositions within a larger context de ned rst in [Karttunen, 1974] and ....

Knott, A. (1996). A Data-driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, Department of Articial Intelligence, University of Edinburgh.

....tree for the presentational relations appears in gure 1. When marking these relations, remember to follow the tree, and at the end con rm your choice by checking the relation de nition and the examples. The de nitions for these relations come from [4] The discourse cues come from [4] and [2]. Antithesis Conditions: 1. The situations of N and S are in contrast. 2. One of the di erences between N and S means that N and S can t simultaneously be considered to be true, valid, good or preferable. 3. The speaker views the situation of N as true, valid, good or preferable. 4. Knowing S, ....

....The decision tree for these relations starts in gure 2. When marking these relations, remember to follow the tree, and at the end con rm your choice by checking the relation de nition and the examples. Most of the de nitions for these relations come from [4] The discourse cues come from [4] and [2]. 15 Abstract:instance Conditions: 1. S presents an instance of, or example of, the concept described in N. Contrast with: set:member. Cues: for instance , for example Example 1 (N) The pongid family of the order primate is closely related to mankind. S) A particularly interesting ....

[Article contains additional citation context not shown here]

A. Knott. A Data-Driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, Department of Articial Intelligence, University of Edinburgh, 1996.

....fNarr; Elabg fNarr; Explg fElab; Explg fResg fNarrg fElabg fExplg ; Figure 4: The discourse relation lattice for four discourse relations 9 The number of used discourse relations is currently being further investigated. A starting point for this investigation is the work on discourse clues by Knott (1996). Robust Text Analysis via Underspecification 3.1.2. Topic information The topic node plays an important role for the discourse representation. It contains information in an abstract form as to what the given segment is about. Note that the usage of the term topic has varied widely in the ....

KNOTT A. (1996). A Data-Driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, Department of Artificial Intelligence, University of Edinburgh.

....text: Zovirax has no significant side effects; however, the medicine is for you only; consequently, never give Zovirax to other patients. 2. 6 Discourse markers A complete treatment of discourse markers will not be attempted here, but we will cover the three main categories identified by Knott (1996): subordinating conjunctions (subord conj) coordinating conjunctions (coord conj) and conjunctive adverbs (conj adv) The properties of a discourse marker can be fully specified by four features: MEANING, SYNTAX, LOCUS, and SPELLING. As examples, here are three definitions for the concession ....

Knott, A. (1996). A data-driven methodology for motivating a set of coherence relations. Technical report, University of Edinburgh. Ph.D. thesis.

....Hendriks, Jackendo , Krifka, Partee, Peregrin, Rooth, van der Sandt, and Steedman. Our aim in the present paper is to extend the repertoire of IS sensitive accounts in a new direction, building on research that indicates that the meaning of some discourse connectives involves presuppositions [7, 8, 11, 22, 24, 25]. In particular: Negative polarity causal connectives like although , however and nevertheless are taken to presuppose a defeasible rule (PDR) Connectives like otherwise are taken to presuppose one of their arguments. With otherwise , the presupposed argument provides a basis 2 ....

....that the construction appeals to but does not explicitly assert. Lagerwerf [11] was the rst to treat such expectations, which he associated with defeasible rules, as being presuppositions of the relevant discourse connectives. A similar treatment appeared independently in work by Knott [7] and by Knott and Mellish [8] The implications of such a treatment have been discussed by Webber and Knott and their colleagues in [23, 24, 25] According to Webber et al. the arguments of the relation asserted by a discourse connective may be given structurally (as with subordinate clause ....

Alistair Knott. A Data-driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, Department of Articial Intelligence, University of Edinburgh, 1996.

.... In contrast to DRS based semantics traditionally given to phrases of the syntactic type as appearing above, we do not consider lexical items as functions over DRS 1 It should be noted, however, that there are phrases of different type that function as operators on discourse referents ([Kno96]) Such phrases are not discussed in this paper (but cf. LGN99] 2 Of course, the primitive notions Flight, Location, and Direction have to be based on an ontology like that of WordNet. 4 An Inference Based Approach to the Interpretation of Discourse and discourse referents and we assign a ....

A. Knott, A Data-Driven Methodology for Motivating a Set of Coherence Relations, PhD thesis, University of Edinburgh 1996

....sequences have a well defined meaning in the domain model. 3.2 Function Words In this section, we outline an approach how to define semantics for some classes of function words which can be integrated into the already existing semantic framework sketched in the previous section. According to [Kno96], speakers use function words to express coherence between a set of preconditions and the currently uttered proposition. Coherence is based on a defeasible implication rule describing the speaker s assumption about how the arguments of the function words cohere logically. A defeasible implication ....

....possibility to express the defeasible character of the function word s semantics as proposed by Knott. Additionally, the verification whether the assumption holds or not in the current dialogue can be delayed until the utterance is integrated into the dialogue context (see the following section) [Kno96] also discusses a number of additional semantic properties for function words which in particular serve for describing the relation of an utterance to the discourse (i.e. to what has been said so far) and to the assumptions the speaker makes about the domain. For a detailed discussion of these ....

[Article contains additional citation context not shown here]

A. Knott, A Data-Driven Methodology for Motivating a Set of Coherence Relations, PhD thesis, University of Edinburgh 1996

.... et al. 1992 ] it clefts [ Delin and Oberlander, 1992 ] and discourse markers or cue phrases [ Ballard et al. 1971; Halliday and Hasan, 1976; van Dijk, 1979; Longacre, 1983; Grosz and Sidner, 1986; Schiffrin, 1987; Cohen, 1987; Redeker, 1990; Sanders et al. 1992; Hirschberg and Litman, 1993; Knott, 1995; Fraser, 1996; Moser and Moore, 1997 ] see [ Marcu, 1997 ] for details) Intentional judgments can be obtained by applying the shared plans framework [ Grosz and Kraus, 1996 ] in the style of Lochbaum [ 1998 ] or by exploiting patterns of pronominalization and anaphoric usages [ Sidner, 1981; ....

.... et al. 1995 ] it clefts [ Delin and Oberlander, 1992 ] and discourse markers or cue phrases [ Ballard et al. 1971; Halliday and Hasan, 1976; van Dijk, 1979; Longacre, 1983; Grosz and Sidner, 1986; Schiffrin, 1987; Cohen, 1987; Redeker, 1990; Sanders et al. 1992; Hirschberg and Litman, 1993; Knott, 1995; Fraser, 1996; Moser and Moore, 1997; Marcu, 1997 ] But once these judgments are made, we still need to determine all discourse interpretations that are not only consistent with these judgments but also valid. This paper does not provide semantics for discourse relations and does not explain ....

Alistair Knott. A Data-Driven Methodology for Motivating a Set of Coherence Relations. PhD thesis, University of Edinburgh, 1995.

....can later be integrated with (or compiled into) an account of speaker intentions. With respect to teminology, the discourse connectives of the title also go under such other names as discourse markers (Schi rin, 1987) clausal connectives (Knott and Mellish, 1996) discourse cues, cue phrases (Knott, 1996) and clue phrases (Cohen, 1987) They comprise words and phrases whose use requires an on going discourse (i.e. the presence of at least one clause other than that to which the connective is attached) and whose meaning involves (in part) that discourse. The particular phenomenon considered here ....

....can be stretched long distance, as in Apples, Bill thinks John may like. In discourse, local dependencies can be stretched long distance as well for example 2 While Webber and Joshi (1998) discuss reasons for taking the lexical anchors of these initial trees to be, following the analysis in (Knott, 1996; Knott and Mellish, 1996) feature structures that may correspond to one or more subordinate conjunctions such as if and when , here we just take them to be speci c lexical items. iwcs klu.tex; 17 10 2000; 12:06; p.4 5 D c D c D c and, or but, so . DConn DConn (b) a) ....

[Article contains additional citation context not shown here]

Knott, A.: 1996, `A Data-driven Methodology for Motivating a Set of Coherence Relations'. Ph.D. thesis, Department of Articial Intelligence, University of Edinburgh.

....Relations: A Structural and Presuppositional Account Using Lexicalised TAG Bonnie Webber Univ of Edinburgh bonnie dai.ed. ac.uk Alistair Knott Univ of Otago alik cs.otago.ac.nz Matthew Stone Rutgers Univ mdstone cs.rutgers.edu Aravind Joshi Univ of Pennsylvania joshi cis.upenn.edu Abstract We show that discourse structure need not bear the full burden of conveying discourse relations by showing that many of them can be ....

Alistair Knott. 1996. A Data-driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, Department of Artificial Intelligence, University of Edinburgh.

....relations are associated with particular classes of linguistic expression. The present authors have argued that the set of linguistic resources available for signalling relations in a language can provide valuable evidence for determining how the set of relations in that language should be de ned (Knott, 1996; Knott and Mellish, 1996) and that the lack of conjunctive signals for elaboration provides evidence that it is di erent from other relations. 7 5 An elaborationless model of text coherence While many of the problems with elaboration have been noted in the past, the question of what an ....

Knott, A. (1996). A Data-Driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, Department of Articial Intelligence, University of Edinburgh.

No context found.

Knott, A.: 1996. A Data-Driven Methodology for Motivating a Set of Coherence Relations. PhD Thesis. University of Edinburgh, Edinburgh.

No context found.

Knott, Alistair (1996). A Data-Driven Methodology for Motivating a Set of Coherence Relations, (Ph.D. thesis). Edinburgh: University of Edinburgh.

No context found.

Alistair Knott. 1996. A Data-driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, University of Edinburgh.

No context found.

Alistair Knott. 1996. A Data-Driven Methodology for Motivating a Set of Coherence Relations. Ph.d. thesis, University of Edinburgh.

No context found.

Knott, A. (1996) A Data-driven Methodology for Motivating a Set of Coherence Relations, Phd, University of Edinburgh.

No context found.

Alistair Knott. 1996. A Data-Driven Methodology for Motivating a Set of Coherence Relations.

No context found.

Alistair Knott. 1995. A Data-Driven Methodology for Motivating a Set of Coherence Relations. Ph.D. thesis, University of Edinburgh.

No context found.

"A Data-Driven Methodology for Motivating a Set of Coherence Relations", PhD Thesis, University of Edinburgh

First 50 documents

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