A Text Relatedness and Dependency Computational Model

نویسندگان

  • Cédrick Bellissens
  • Patrick Jeuniaux
  • Nicholas D. Duran
  • Danielle S. McNamara
چکیده

The Interactive Strategy Trainer for Active Reading and Thinking (iSTART) is an intelligent tutoring system that provides students with automated training on reading strategies. In particular, iSTART helps students integrate textual information into a coherent mental representation through self-explanation. The goal of the present study was to examine how text cohesion influences qualitatively different types of self-explanation, namely, bridging and elaborative inferences. To do so, we developed a computational model that characterizes cohesion in terms of the textbase indices word stem and Latent Semantic Analysis relatedness, as well as the situation model index causal dependency between sentences. This model successfully predicted the different types of self-explanations as a function of cohesion. We also found that students’ prior knowledge interacted particularly with causal dependency. RÉSUMÉ. ISTART (Interactive Strategy Trainer for Active Reading and Thinking) est un tuteur électronique qui entraine des étudiants à utiliser certaines stratégies de lecture dans le but de comprendre un texte difficile. Les utilisateurs d’iSTART sont amenés à intégrer les Studia Informatica Universalis. informations, contenues dans un texte, à une représentation mentale cohérente, par la technique de l’auto-explication. Le but de notre étude était d’examiner comment la cohésion textuelle influençait qualitativement la production de différents types d’auto-explications, particulièrement les stratégies consistant à expliciter des liens entre phrases ou à élaborer le contenu d’un texte. Dans cette optique, nous avons développé un modèle qui calcule deux genres de cohésion textuelle à partir d’indices linguistiques différents contribuant, d’une part, à la construction de la base de texte, à savoir, la répétition de racines de mots et la similarité sémantique entre phrases (Analyse Sémantique Latente), et d’autre part, à la construction d’un modèle de situation, comme la dépendance causale. Ce modèle permet de prédire différents types d’auto-explications en fonction du degré de cohésion textuelle calculée. Nous montrons en outre que les connaissances initiales interagissent particulièrement avec la cohésion situationnelle, calculée à partir de la dépendance causale entre phrases.

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

Using Dependency Parses to Augment Feature Construction for Text Mining

(ABSTRACT) With the prevalence of large data stored in the cloud, including unstructured information in the form of text, there is now an increased emphasis on text mining. A broad range of techniques are now used for text mining, including algorithms adapted from machine learning, NLP, computational linguistics, and data mining. Applications are also multi-fold, including classification, clust...

متن کامل

Semantic Relatedness: Computational Investigation of Human Data

We develop computational measures of semantic relatedness using data collected in psycholinguistic and annotation-style experiments. Datasets and measures are briefly described, and correlations with human data are presented. Estimating the degree of semantic relatedness between words in a text is deemed important in numerous applications: word-sense disambiguation (Banerjee and Pedersen, 2003)...

متن کامل

Joint Unsupervised Learning of Semantic Representation of Words and Roles in Dependency Trees

In this paper, we introduce WoRel, a model that jointly learns word embeddings and a semantic representation of word relations. The model learns from plain text sentences and their dependency parse trees. The word embeddings produced by WoRel outperform Skip-Gram and GloVe in word similarity and syntactical word analogy tasks and have comparable results on word relatedness and semantic word ana...

متن کامل

Semantic Relatedness Using Salient Semantic Analysis

This paper introduces a novel method for measuring semantic relatedness using semantic profiles constructed from salient encyclopedic features. The model is built on the notion that the meaning of a word can be characterized by the salient concepts found in its immediate context. In addition to being computationally efficient, the new model has superior performance and remarkable consistency wh...

متن کامل

A New Method for Improving Computational Cost of Open Information Extraction Systems Using Log-Linear Model

Information extraction (IE) is a process of automatically providing a structured representation from an unstructured or semi-structured text. It is a long-standing challenge in natural language processing (NLP) which has been intensified by the increased volume of information and heterogeneity, and non-structured form of it. One of the core information extraction tasks is relation extraction wh...

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:
  • Stud. Inform. Univ.

دوره 8  شماره 

صفحات  -

تاریخ انتشار 2010