Semantic Analysis of Winograd Schema No. 1

The Winograd Schema Challenge is a general test for Artificial Intelligence, based on problems of pronoun reference resolution. I investigate the semantics and interpretation of Winograd Schemas, concentrating on the original and most famous example. This study suggests that a rich ontology, detailed commonsense knowledge as well as special purpose inference mechanisms are all required to resolve just this one example. The analysis supports the view that a key factor in the interpretation and disambiguation of natural language is the preference for coherence. This preference guides the resolution of co-reference in relation to both explicitly mentioned entities and also implicit entities that are required to form an interpretation of what is being described. I suggest that assumed identity of implicit entities arises from the expectation of coherence and provides a key mechanism that underpins natural language understanding. I also argue that conceptual ontologies can play a decisive role not only in directly determining pronoun references but also in identifying implicit entities and implied relationships that bind together components of a sentence.

Critical Scholarship on Terrorism

Understandings of “critical” in critical scholarship on terrorism range from a Frankfurt School–influenced definition to a broader definition that aims to interrogate commonsense understandings of terrorism and counterterrorism. Overall, critical scholarship on terrorism draws on multiple disciplines and methodological traditions to analyze terrorism and counterterrorism. Within these, there have been ongoing debates and discussions about whether the state should be included in research on terrorism and, if so, what the inclusion of the state would do for the understanding of terrorism. Critical scholarship has also outlined the need for further attention to research ethics, as well as urged researchers to acknowledge their standpoints when conducting and communicating research. Some, but not all, critical scholarship has a normative orientation with the goal of emancipation, though the meaning of emancipation remains debated. Methodologically, the majority of critical scholarship on terrorism utilizes an interpretive lens to analyze terrorism and related issues. A central goal of critical terrorism research is to rework power relations such that Global South subjectivities are centered on research. This means including research conducted by Global South scholars and also centering Global South peoples and concerns in analyses of terrorism and counterterrorism. The role of gender, analytically and in practice, in relation to terrorism is also a key part of critical scholarship. Critical scholars of terrorism have observed that race is absent from much of terrorism scholarship, and there needs to be ongoing work toward addressing this imbalance. Media and popular culture, and their depiction of terrorism and counterterrorism, form another key strand in critical scholarship on terrorism. Overall, critical scholarship on terrorism is about scrutinizing and dismantling power structures that sustain commonsense knowledge regarding terrorism.

Queries and Cues: Textual Stimuli for Reflective Thinking in Digital Mind-Mapping

Mind-mapping is useful for externalizing ideas and their relationships surrounding a central problem. However, balancing between the exploration of different aspects (breadth) of the problem with respect to the detailed exploration of each of its aspects (depth) can be challenging, especially for novices. The goal of this paper is to investigate the notion of “reflection-in-design” through a novel interactive digital mind-mapping workflow that we call “QCue”. The idea behind this workflow is to incorporate the notion of reflective thinking through two mechanisms: (1) offering suggestions to promote depth exploration through user's queries (Q), and (2) asking questions (Cue) to promote reflection for breadth exploration. This paper is an extension of our prior work where our focus was mainly on the algorithmic development and implementation of a cognitive support mechanism behind QCue enabled by ConceptNet (a graph-based rich ontology with “commonsense” knowledge). In this extended work, we first present a detailed summary of how QCue facilitated the breadth-depth balance in a mind-mapping task. Second, we present a comparison between QCue and conventional digital mind-mapping i.e. without our algorithm through a between-subjects user study. Third, we present new detailed analysis on the usage of different cognitive mechanisms provided by QCue. We further consolidate our prior quantitative analysis and build a connection with our observational analysis. Finally, we discuss in detail the different cognitive mechanisms provided by QCue to stimulate reflection in design.

PATCHCOMM: Using Commonsense Knowledge to Guide Syntactic Parsers

Syntactic parsing technologies have become significantly more robust thanks to advancements in their underlying statistical and Deep Neural Network (DNN) techniques: most modern syntactic parsers can produce a syntactic parse tree for almost any sentence, including ones that may not be strictly grammatical. Despite improved robustness, such parsers still do not reflect the alternatives in parsing that are intrinsic in syntactic ambiguities. Two most notable such ambiguities are prepositional phrase (PP) attachment ambiguities and pronoun coreference ambiguities. In this paper, we discuss PatchComm, which uses commonsense knowledge to help resolve both kinds of ambiguities. To the best of our knowledge, we are the first to propose the general-purpose approach of using external commonsense knowledge bases to guide syntactic parsers. We evaluated PatchComm against the state-of-the-art (SOTA) spaCy parser on a PP attachment task and against the SOTA NeuralCoref module on a coreference task. Results show that PatchComm is successful at detecting syntactic ambiguities and using commonsense knowledge to help resolve them.

Strong Influence of Responses in Training Dialogue Response Generator

The sequence-to-sequence model is a widely used model for dialogue response generators, but it tends to generate safe responses for most input queries. Since safe responses are unattractive and boring, a number of efforts have been made to make the generator produce diverse responses, but generating diverse responses is yet an open problem. As a solution to this problem, this paper proposes a novel response generator, Response Generator with Response Weight (RGRW). The proposed response generator is a transformer-based sequence-to-sequence model of which the encoder is a pre-trained Bidirectional Encoder Representations from Transformers (BERT) and the decoder is a variant of Generative Pre-Training of a language model-2 (GPT-2). Since the attention on the response is not reflected enough at the transformer-based sequence-to-sequence model, the proposed generator enhances the influence of a response by the response weight, which determines the importance of each token in a query with respect to the response. Then, the decoder of the generator processes the response weight as well as a query encoding to generate a diverse response. The effectiveness of RGRW is proven by showing that it generates more diverse and informative responses than the baseline response generator by focusing more on the tokens that are important for generating the response. Additionally, the proposed model overwhelms the Commonsense Knowledge-Aware Dialogue generation model (ConKADI), which is a state-of-the-art model.

TANGO: Commonsense Generalization in Predicting Tool Interactions for Mobile Manipulators

Robots assisting us in factories or homes must learn to make use of objects as tools to perform tasks, e.g., a tray for carrying objects. We consider the problem of learning commonsense knowledge of when a tool may be useful and how its use may be composed with other tools to accomplish a high-level task instructed by a human. We introduce TANGO, a novel neural model for predicting task-specific tool interactions. TANGO is trained using demonstrations obtained from human teachers instructing a virtual robot in a physics simulator. TANGO encodes the world state consisting of objects and symbolic relationships between them using a graph neural network. The model learns to attend over the scene using knowledge of the goal and the action history, finally decoding the symbolic action to execute. Crucially, we address generalization to unseen environments where some known tools are missing, but alternative unseen tools are present. We show that by augmenting the representation of the environment with pre-trained embeddings derived from a knowledge-base, the model can generalize effectively to novel environments. Experimental results show a 60.5-78.9% improvement over the baseline in predicting successful symbolic plans in unseen settings for a simulated mobile manipulator.