Mark Hopkins

What's in a Translation Rule

Abstract : We propose a theory that gives formal semantics to word-level alignments defined over parallel corpora. We use our theory to introduce a linear algorithm that can be used to derive from word-aligned, parallel corpora the minimal set of syntactically motivated transformation rules that explain human translation data.

The Power of Extended Top-Down Tree Transducers

Extended top-down tree transducers (transducteurs generalises descendants; see [A. Arnold and M. Dauchet, Bi-transductions de forets, in Proceedings of the 3rd International Colloquium on Automata, Languages and Programming, Edinburgh University Press, Edinburgh, 1976, pp. 74-86]) received renewed interest in the field of natural language processing. Here those transducers are extensively and systematically studied. Their main properties are identified and their relation to classical top-down tree transducers is exactly characterized. The obtained properties completely explain the Hasse diagram of the induced classes of tree transformations. In addition, it is shown that most interesting classes of transformations computed by extended top-down tree transducers are not closed under composition.

Causality and Counterfactuals in the Situation Calculus

Structural causal models offer a popular framework for exploring causal concepts. However, due to their limited expressiveness, structural models have difficulties coping with such concepts as actual (event-to-event) causation. In this article, we propose a new type of causal model, based on embedding structural considerations in the language of situation calculus. By using situation calculus as a basic language, we leverage its power to express complex, dynamically changing situations and, by relying on structural considerations, we can formulate an effective theory of counterfactuals within the situation-calculus.

Cube Pruning as Heuristic Search

Cube pruning is a fast inexact method for generating the items of a beam decoder. In this paper, we show that cube pruning is essentially equivalent to A* search on a specific search space with specific heuristics. We use this insight to develop faster and exact variants of cube pruning.

Machine translation as tree labeling

We present the main ideas behind a new syntax-based machine translation system, based on reducing the machine translation task to a tree-labeling task. This tree labeling is further reduced to a sequence of decisions (of four varieties), which can be discriminatively trained. The optimal tree labeling (i.e. translation) is then found through a simple depth-first branch-andbound search. An early system founded on these ideas has been shown to be competitive with Pharaoh when both are trained on a small subsection of the Europarl corpus.

A framework for incorporating alignment information in parsing

The standard PCFG approach to parsing is quite successful on certain domains, but is relatively inflexible in the type of feature information we can include in its probabilistic model. In this work, we discuss preliminary work in developing a new probabilistic parsing model that allows us to easily incorporate many different types of features, including crosslingual information. We show how this model can be used to build a successful parser for a small handmade gold-standard corpus of 188 sentences (in 3 languages) from the Europarl corpus.

Deep Grammars in a Tree Labeling Approach to Syntax-based Statistical Machine Translation

In this paper, we propose a new syntaxbased machine translation (MT) approach based on reducing the MT task to a tree-labeling task, which is further decomposed into a sequence of simple decisions for which discriminative classifiers can be trained. The approach is very flexible and we believe that it is particularly well-suited for exploiting the linguistic knowledge encoded in deep grammars whenever possible, while at the same time taking advantage of data-based techniques that have proven a powerful basis for MT, as recent advances in statistical MT show. A full system using the Lexical-Functional Grammar (LFG) parsing system XLE and the grammars from the Parallel Grammar development project (ParGram; (Butt et al., 2002)) has been implemented, and we present preliminary results on English-to-German translation with a tree-labeling system trained on a small subsection of the Europarl corpus.

Exploring the Potential of Intractable Parsers

We revisit the idea of history-based parsing, and present a history-based parsing framework that strives to be simple, general, and flexible. We also provide a decoder for this probability model that is linear-space, optimal, and anytime. A parser based on this framework, when evaluated on Section 23 of the Penn Tree-bank, compares favorably with other state-of-the-art approaches, in terms of both accuracy and speed.