Choosing Rhetorical Structures To Plan Instructional Texts

This paper discusses a fundamental problem in natural language generation: how to organize the content of a text in a coherent and natural way. In this research, we set out to determine the semantic content and the rhetorical structure of texts and to develop heuristics to perform this process automatically within a text generation framework. The study was performed on a specific language and textual genre: French instructional texts. From a corpus analysis of these texts, we determined nine senses typically communicated in instructional texts and seven rhetorical relations used to present these senses. From this analysis, we then developed a set of presentation heuristics that determine how the senses to be communicated should be organized rhetorically in order to create a coherent and natural text. The heuristics are based on five types of constraints: conceptual, semantic, rhetorical, pragmatic, and intentional constraints. To verify the heuristics, we developed the spin natural language generation system, which performs all steps of text generation but focuses on the determination of the content and the rhetorical structure of the text.     

évolution du trouble d’opposition et du trouble des conduites au cours de l’enfance.

This study aimed at describing the evolution of the symptoms of the oppositional disorder and conducts disorder on a 4-year period (3 times of measurement), depending on whether both disorders appear simultaneously or separately. The sample was composed of 336 children (6–13 years-old), separated in three groups according to the nature of the disorder they presented at the beginning of the study: oppositional disorder only, conducts disorder only, both oppositional and conducts disorders. Structured diagnostic interviews based on the diagnostic criterions of the Diagnostic and Statistical Manual of Mental Disorders (4 éd., texte rév.; American Psychiatric Association, 2000) were used in order to evaluate the average number of symptoms associated to each disorder at each time of measurement. Multilevel regression analysis suggest that the initial co-occurrence of the disorders is neither associated to the evolution of the symptoms of the conducts disorder nor to those of the oppositional disorder. However, this co-occurrence, that can also include the attention-deficit/hyperactivity disorder, can increase the number of symptoms in children, which likely contributes to maintain the diagnostics in some cases. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Generating paraphrases from meaning-text semantic networks

This paper describes a first attempt to base a paraphrase generation system upon Meľčuk and Žolkovskij's linguistic meaning-text (MT) model whose purpose is to establish correspondences between meanings, represented by networks, and (ideally) all synonymous texts having this meaning. The system described here contains a Prolog implementation of a small explanatory and combinatorial dictionary (the MT lexicon) and, using unification and backtracking, generates from a given network the sentences allowed by the dictionary and the lexical transformations of the model. The passage from a net to the final texts is done through a series of transformations of intermediary structures that closely correspond to MT utterance representations (semantic, deep-syntax, surface-syntax, and morphological representations). These are graphs and trees with labeled arcs. The Prolog unification (equality predicate) was extended to extract information from these representations and build new ones. The notion of utterance path, used by many authors, is replaced by that of covering by defining subnetworks.     

Assessment of the method for calculating the Lewis number of H2/CO/CH4 mixtures and comparison with experimental results

This paper investigates the various techniques used in the literature to calculate the effective Lewis number of two-component (H₂/CO and H₂/CH₄) and three-component fuels (H₂/CO/CH₄ and H₂/CO/CO₂) over a wide range of equivalence ratios (0.6 ≤ φ ≤ 1.4) under laminar flame conditions. The most appropriate effective Lewis number formulation is identified through comparison with experimentally extracted Lewis numbers (Le). The paper first identifies the proper methodology to extract the experimental Le from the burned Markstein length of an outwardly propagating flame. Second, the different methodologies for the calculation of the effective Le are presented and compared to experimental results for H₂/CH₄ and H₂/CO mixtures. Based on the experimental results, it is shown that the calculation of the effective Le of mixtures can be divided into a three-step procedure depending on the equivalence ratio: (1) calculation of the Le for each fuel and the oxidizer; (2) use of the Le mixing rule; and (3) assessment of the necessity or not of combining the fuel's and oxidizer's Lewis numbers. The paper shows that, in rich mixtures, the oxidizer Le needs to be taken into account. Lastly, the methodology is validated for H₂/CO/CH₄ and H₂/CO/CO₂ fuels.     

FUS: a system to simulate conformational changes in biological macromolecules

In order to study the dynamics of protein and nucleic acid conformations, a molecular folding-unfolding system (FUS written in Lisp) has been developed. Secondary structure features of protein and nucleic acids are graphically represented by cubes in a modified 'Blocks World' paradigm. Modeling of protein and nucleic acid unfolding (denaturation) and folding of their three-dimensional structure is possible by the use of high level 'block' operators which allow displacement of these structural features in space. Due to the flexible nature of this program, FUS is a useful tool for the rapid evaluation of user-defined rules governing conformational changes. The use of FUS to unfold three common proteins (prealbumin, flavodoxin and triose phosphate isomerase) and a tRNA is presented.     

Unit Completion for a Computer-aided Translation Typing System

This work is in the context of, a system thatwatches over the users as they type a translation andrepeatedly suggests completions for the text already entered.The users may either accept, modify, or ignore these suggestions. Wedescribe the design, implementation, and performance of aprototype which suggests completions of units of texts that arelonger than one word.     

The Long-Term Forecast for Weather Bulletin Translation

Machine Translation (MT) is the focus of extensive scientific investigations driven by regular evaluation campaigns, but which are mostly oriented towards a somewhat particular task: translating news articles into English. In this paper, we investigate how well current MT approaches deal with a real-world task. We have rationally reconstructed one of the only MT systems in daily use which produces high-quality translation: the Météo system. We show how a combination of a sentence-based memory approach, a phrase-based statistical engine and a neural-network rescorer can give results comparable to those of the current system. We also explore another possible prospect for MT technology: the translation of weather alerts, which are currently being translated manually by translators at the Canadian Translation Bureau.     

Influence of CO2, CH4, and initial temperature on H2/CO laminar flame speed

The objective of this study is to investigate the impact of syngas composition by varying the H₂/CO ratio (1:3, 1:1, and 3:1 by volume), the CO₂ dilution (0%–40%), and methane addition (0%–40%) on laminar flame speed. Thus, laminar flame speeds of premixed syngas–air mixtures were measured for different equivalence ratios (0.8–2.2) and inlet temperatures (295–450 K) using the Bunsen-burner method. It was found that laminar flame speed increases with increasing H₂/CO ratio, while CO₂ dilution or CH₄ addition decreased it. The location of the maximum flame speed shifts to richer mixtures with decreasing H₂/CO ratio, while it shifts to leaner mixtures with the addition of CH₄ due to its inherent slower flame speed. The location of the maximum flame speed is also shifted towards leaner mixtures with the addition of CO₂ due to the preponderance of the reduction of the adiabatic flame temperature with increasing dilution. Comparison between experimental and numerical results shows a better agreement using a modified mechanism derived from GRI-Mech 3.0. A correlation, based on the experimental results, is proposed to calculate the laminar flame speed over a wide range of equivalence ratios, inlet temperatures, and fuel content.     

Trans Type: Development-Evaluation Cycles to Boost Translator's Productivity

We present TransType: a new approach to Machine-Aided Translation in which the human translator maintains control of the translation process while being helped by real-time completions proposed by a statistical translation engine. The TransType approach is first presented through a series of prototypes that illustrate their underlying translation model and graphical interface. The results of two rounds of in situ evaluation of TransType prototypes are discussed followed by a set of lessons learned in these experiments. It will be shown that this approach is valued by translators but given the short time allotted for the evaluation, translators were not able to quantitatively increase their productivity. TransType is compared with other approaches and new perspectives are elaborated for a new version being developed in the context of a Fifth Framework European Community Project. This revised version was published online in November 2006 with corrections to the Cover Date.     

Biomagnetic source detection by maximum entropy and graphical models

This article presents a new approach for detecting active sources in the cortex from magnetic field measurements on the scalp in magnetoencephalography (MEG). The solution of this ill-posed inverse problem is addressed within the framework of maximum entropy on the mean (MEM) principle introduced by Clarke and Janday. The main ingredient of this regularization technique is a reference probability measure on the random variables of interest. These variables are the intensity of current sources distributed on the cortical surface for which this measure encompasses all available prior information that could help to regularize the inverse problem. This measure introduces hidden Markov random variables associated with the activation state of predefined cortical regions. MEM approach is applied within this particular probabilistic framework and simulations show that the present methodology leads to a practical detection of cerebral activity from MEG data.