The Story Cloze Test consists of choosing a sentence that best completes a story given two choices. In this paper we present a system that performs this task using a supervised binary classifier on top of a recurrent neural network to predict the probability that a given story ending is correct. The classifier is trained to distinguish correct story endings given in the training data from incor...

We present Guru, an intelligent tutoring system for high school biology that has conversations with students, gestures and points to virtual instructional materials, and presents exercises for extended practice. Guru’s instructional strategies are modeled after expert tutors and focus on brief interactive lectures followed by rounds of scaffolding as well as summarizing, concept mapping, and Cl...

Idioms and common multi-word expressions are often argued to be stored as chunks of words or fixed configurations in the mind, and to therefore be accessed faster and interpreted more easily than fully compositional word combinations. Experimental research has furthermore shown that a specific “recognition point” can be identified in such expressions, at which enough information is present to a...

Semantic scripts is a conceptual representation which defines how events are organized into higher level activities. Practically all the previous approaches to inducing script knowledge from text relied on count-based techniques (e.g., generative models) and have not attempted to compositionally model events. In this work, we introduce a neural network model which relies on distributed composit...

We present two NLP components for the Story Cloze Task – dictionary-based sentiment analysis and lexical cohesion. While previous research found no contribution from sentiment analysis to the accuracy on this task, we demonstrate that sentiment is an important aspect. We describe a new approach, using a rule that estimates sentiment congruence in a story. Our sentiment-based system achieves str...

Scripts represent knowledge of stereotypical event sequences that can aid text understanding. Initial statistical methods have been developed to learn probabilistic scripts from raw text corpora; however, they utilize a very impoverished representation of events, consisting of a verb and one dependent argument. We present a script learning approach that employs events with multiple arguments. U...

This paper reports on an investigation of lexical selfrepair in Dutch spontaneous dialogue. Lexical selfrepairs, in which one word is rejected for another, can be produced with or without notable ̳prosodic marking‘ of the second word. It remains unclear what motivates speakers‘ choices, but previous research has shown that the semantic distance between the two words is relevant. This study asse...

Implicit arguments are not syntactically connected to their predicates, and are therefore hard to extract. Previous work has used models with large numbers of features, evaluated on very small datasets. We propose to train models for implicit argument prediction on a simple cloze task, for which data can be generated automatically at scale. This allows us to use a neural model, which draws on n...

Two large-scale cloze-style context-question-answer datasets have been introduced recently: i) the CNN and Daily Mail news data and ii) the Children’s Book Test. Thanks to the size of these datasets, the associated task is well suited for deep-learning techniques that seem to outperform all alternative approaches. We present a new, simple model that is tailor made for such question-answering pr...

We present the task of answering cloze questions automatically and how it can be tackled by exploiting lexical knowledge bases (LKBs). This task was performed in what can be seen as an indirect evaluation of Portuguese LKB. We introduce the LKBs used and the algorithms applied, and then report on the obtained results and draw some conclusions: LKBs are definitely useful resources for this chall...