Predicting Liaison: an Example-Based Approach

نویسندگان

  • Antal van den Bosch
  • Alexander Greefhorst
چکیده

Predicting liaison in French is a non-trivial problem to model. We compare a memory-based machine-learning algorithm with a rule-based baseline. The memory-based learner is trained to predict whether liaison occurs between two words on the basis of lexical, orthographic, morphosyntactic, and sociolinguistic features. Best performance is obtained using only a selection of lexical and syntactic features, yielding a best overall performance at a precision of .80, with recall at .85. Counter to our expectations, including sociolinguistic features even lowered the precision and recall of our predictions. The F-scores of the memory-based algorithm are higher than that of a simple baseline and three other state-ofthe-art machine-learning algorithms. Based on the results on optional liaison, it appears that predicting liaison benefits from being able to generalize from specific examples in context. RÉSUMÉ. La prédiction de la liaison en français est un problème de modélisation non trivial. Nous comparons un algorithme d’apprentissage automatique basé sur la mémoire avec un point de comparaison basé sur des règles. L’apprentissage automatique est entraîné à prédire si la liaison se produit entre deux mots consécutifs en évaluant des traits lexicaux, orthographiques, morphosyntaxiques et sociolinguistiques. Notre étude montre que la meilleure performance est obtenue en utilisant uniquement des traits lexicaux et syntaxiques, résultant en une précision de .80 et un rappel de .85. Contrairement à nos attentes, l’inclusion des traits sociolinguistiques rend la précision et le rappel plus bas. La F-mesure est la plus élevée en utilisant l’algorithme d’apprentissage automatique basé sur la mémoire. Elle est non seulement plus élevée que le point de comparaison basé sur des règles, mais aussi plus élevée que celle de trois autres algorithmes d’apprentissage automatique de pointe. Il paraît que la possibilité de généralisation des exemples spécifiques en contexte aide la prédiction de la liaison.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Predicting Customer-Expectation-Based Warranty Cost for Smaller-the- Better and Larger-the-Better Performance Characteristics

The quality loss function assumes a fixed target and only accounts for immediate issues within manufacturing facilities whereas warranty loss occurs during customer use. Based on the two independent variables, product performance and consumers’ expectation, a methodology to predict the probability of customer complaint is presented in this paper. The formulation presented will serve as a basic ...

متن کامل

Predicting ε50 for Lateral Behavior of Piles in Marine Clay Using an Evolutionary Based Approach

Analyzing piles subjected to lateral loads significantly depends on soil resistance at any point along the pile as a function of pile deflection, known as p-y curve. On the other hand, the deformation characteristics of soil defined as “the soil strain at 50% of maximum deviatoric stress (ε50)” has considerable effect on the generated p-y curve. In this research, several models are proposed to ...

متن کامل

An ANFIS-based Approach for Predicting the Manning Roughness Coefficient in Alluvial Channels at the Bank-full Stage

An intelligent method based on adaptive neuro-fuzzy inference system (ANFIS) for identifying Manning’s roughness coefficients in modeling of alluvial river is presented. The procedure for selecting values of Manning n is subjective and requires judgment and skill which are developed primarily through experience. During practical applications, researchers often find that a correct choice of the ...

متن کامل

Evaluating the Role of Company Life Cycle for an Appropriate Model in Predicting the Quality of Discretionary Accruals (Abnormal) Based on Dickinson Cash Flow Model Approach

The main purpose of this research is to evaluate the role of the company life cycle in providing an appropriate model in predicting the quality of discretionary accruals (Abnormal) using the Dickinson Cash Flow Model approach. The statistical population of the research consisted of 180 company observations that were divided into three stages of life cycle using Dickinson's model variables (2011...

متن کامل

Bridging the gap of domain and visualization experts with a Liaison

We introduce the role Liaison for design study projects. With considerable expertise in visualization and the application domain, a Liaison can help to foster richer and more effective interdisciplinary communication in problem characterization, design, and evaluation processes. We characterize this role, provide a list of tasks of Liaison and visualization experts, and discuss concrete benefit...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:
  • TAL

دوره 57  شماره 

صفحات  -

تاریخ انتشار 2016