In this paper, a system is described that can understand a plot of a story on reference to both image and linguistic information. As input, a series of line drawings with colors and narrations in English concerning to these drawings are given to the system. It searches the objects suggested to be in the scene by the narrations, finding relations among them, making the world model by using its w...

One of the main challenges in the formal modeling of common-sense reasoning is the ability to cope with the dynamic nature of the world. Among the approaches put forward to address this problem are belief revision and update. Given a knowledge base T , representing our knowledge of the “state of affairs” of the world of interest, we are often led to trust another piece of information P , possib...

Most sentiment analysis work has been carried out on highly subjective text types where the target is clearly defined and unique across the text (movie or product reviews). However, when applying sentiment analysis to the news domain, it is necessary to clearly define the scope of the task, in a more specific manner than it has been done in the field so far. The main tasks we identified for new...

This paper presents a nonmonotonic deontic logic based on commonsense entailment. It establishes criteria a successful account of obligation should satisfy, and develops a theory that satisfies them. The theory includes two conditional notions of prima facie obligation. One is constitutive; the other is epistemic, and follows nonmonotonically from the constitutive notion. The paper defines unco...

This work presents a novel approach for human action recognition based on the combination of computer vision techniques and common-sense knowledge and reasoning capabilities. The emphasis of this work is on how common sense has to be leveraged to a vision-based human action recognition so that nonsensical errors can be amended at the understanding stage. The proposed framework is to be deployed...

Noise-resistant and nuanced, COGBASE makes 10 million pieces of commonsense data and a host of novel reasoning algorithms available via a family of semantically-driven prior probability distributions. Machine learning, Big Data, natural language understanding/processing, and social AI can draw on COGBASE to determine lexical semantics, infer goals and interests, simulate emotion and affect, cal...

In this report we present the work done on the first subtask of the DEFT 2014 challenge which dealt with genre classification of French literary texts. In our approach we developed three types of features : lemmatized words, stylometric features and features that incorporate some form of world knowledge. Subsequent classification experiments were performed using the Balanced Winnow classifier. ...

Recent theories of concepts have raised the issue of how people combine simple concepts (like engine and repalr) to form complex concepts (like engine repalr). This article approaches this issue by asking how people comprehend modified noun phrases of this sort. One explanation of how complex concepts are understood (the feature weighting model) provides a simple mechanism in which the primary ...

“Better break your word than do worse in keeping it.” The suggestion is particularly compelling when breaking the promise leaves no one worse off, while keeping it is strictly worse for any or all parties. What is the point of performing such a promise when it can be avoided? Granted, dynamic considerations may call for credible commitments to carry out promises that are inconvenient to all con...

Reasoning about motion is an important part of common sense knowledge . The spatial and qualitative aspects of reasoning about motion through free space are studied through the construction of a program to perform such reasoning. An analog geometry representation serves as a diagram, and descriptions of both the actual motion of a ball and envisioning are used in answering simple questions.