Briscoe, T. and Copestake, A. Lexical Rules in Constraint-Based Grammar. Computational Linguistics, v. 25, n. 4, p. 487526, 1999. 213  Home/Search   Document Details and Download   Summary   Related Articles   Check

....uncertain about the thematic roles. Faced with the input data provided by the annotated Sachs corpus, the learner has to set the parameters of the UG accordingly to be able to successfully process this data. The learning algorithm implements a Bayesian Incremental Parameter Setting (BIPS) learner [Briscoe 1999], and when setting the parameters it uses a Minimum Description Length (MDL) style bias to choose the most probable grammar that describes the data well. Several experiments are conducted for testing the performance of the learner in di#erent environments, and the results obtained are analysed to ....

....systems are prone to errors, being able to correctly assign semantic and syntactic categories to only a small portion of the corpus. This is followed by a description, in chapter 6, of the learning system developed. The learning system implements a Bayesian Incremental Parameter Setting algorithm [Briscoe 1999], which uses an MDL style bias to set the parameters of the grammar. The original algorithm proposed by Briscoe is adapted for learning within a unification based framework. The learning system needs to be able to learn new and more complex categories, and include them in the currently used ....

[Article contains additional citation context not shown here]

Briscoe, T. and Copestake, A. Lexical Rules in Constraint-Based Grammar. Computational Linguistics, v. 25, n. 4, p. 487526, 1999. 213

....level, the effect that this rule is supposed to have is clear: anything in the grammar that corresponds to the AVM on the left hand side of the rule should get related to something that corresponds to the AVM on the right hand side. So why is this intuitive understanding not sufficient 10 Briscoe Copestake (1999) provide an interesting discussion of lexical rules in a typed default feature structure framework (Lascarides et al. 1996, Lascarides Copestake 1999) which is an extension of a Kasper Rounds logic (Rounds Kasper 1986, Moshier Rounds 1987, Carpenter 1992) The ontological assumptions and ....

Briscoe, T. & Copestake, A., 1999. Lexical Rules in Constraint-based Grammars. Computational Linguistics 25(4), 487--526. http://www-csli.stanford.edu/aac/papers/lr2-0.ps.gz.

....I think it is useful to consider these phenomena together, because this sheds light on some interesting commonalities and distinctions. I will also describe a possible formal approach. This discussion follows earlier work, in particular the account of productivity in lexical rules proposed in Briscoe and Copestake (1999), but in this paper I consider whether this sort of approach can be used in cases where lexical rules aren t applicable. 2 Semi productivity in alternations and sense extension I take as a starting point the assumption that lexical rules, implemented for instance within a typed feature ....

....grounds to allow such uses to be generated even though they may not have been seen before by most native speakers. Similarly, lexical rules provide a way of automatically allowing for some of the unseen usages of words in a computational implementation. However, as discussed in detail in Briscoe and Copestake (1999) (henceforth B C) while there have been many attempts to define narrow classes within which alternations such as dative are fully productive (e.g. Pinker, 1989) it appears that even though the semantic criteria invoked may be very subtle (and di#cult to test or motivate independently) ....

[Article contains additional citation context not shown here]

Briscoe, Edward J. and Ann Copestake (1999) `Lexical rules in constraint-based grammars', Computational Linguistics, 25:4, 487--526.

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