Analogical Reasoning on Chinese Morphological and Semantic Relations

arXiv.org Artificial Intelligence

Analogical reasoning is effective in capturing linguistic regularities. This paper proposes an analogical reasoning task on Chinese. After delving into Chinese lexical knowledge, we sketch 68 implicit morphological relations and 28 explicit semantic relations. A big and balanced dataset CA8 is then built for this task, including 17813 questions. Furthermore, we systematically explore the influences of vector representations, context features, and corpora on analogical reasoning. With the experiments, CA8 is proved to be a reliable benchmark for evaluating Chinese word embeddings.


Snasci Reasoning

#artificialintelligence

Some of the above list may not be traditionally thought of, or referred to, as types of reasoning. It is only when approaching the practical development of an Artificial General Intelligence is it observed that each of these are approaches or methodologies that constitute a form of reasoning independent of the others. Leveraging these reasoning types and approaches, Snasci will also be quite creative, capable of lying, story telling, humour and adapting to new situations without prior training. In business and scientific applications, Snasci's reasoning and comprehension capabilities will become invaluable. In addition, the ability to connect a wide range of sensors, novel inputs and outputs (including HPC rendering) means that any lab, research and development department, university, etc., can have a world class installation and knowledge base for a fraction of what it currently costs.


Preface

AAAI Conferences

Welcome to the Workshop on Spatial and Temporal Reasoning at AAAI-07 in Vancouver, British Columbia. This workshop continues in the spirit of a series of such activities over the last fifteen years spanning related communities of researchers that study representing and reasoning about either space or time or both. In addition, the workshop has encouraged a mix of theory and applied work. Various basic representational problems in space (direction, location, proximity, geometry, intersection) and in time (coincidence, order, concurrency, overlap, granularity) attract repeated attention due to their fundamental and difficult nature. Beyond that, however, the richness of different ontologies, different applications, and different objectives assures that no small collection of solutions will serve to satisfy all needs.


Forbus

AAAI Conferences

We believe that the flexibility and robustness of common sense reasoning comes from analogical reasoning, learning, and generalization operating over massive amounts of experience. Million-fact knowledge bases are a good starting point, but are likely to be orders of magnitude smaller, in terms of ground facts, than will be needed to achieve human-like common sense reasoning. This paper describes the FIRE reasoning engine which we have built to experiment with this approach. We discuss its knowledge base organization, including coarse-coding via mentions and a persistent TMS to achieve efficient retrieval while respecting the logical environment formed by contexts and their relationships in the KB. We describe its stratified reasoning organization, which supports both reflexive reasoning (Ask, Query) and deliberative reasoning (Solve, HTN planner). Analogical reasoning, learning, and generalization are supported as part of reflexive reasoning. To show the utility of these ideas, we describe how they are used in the Companion cognitive architecture, which has been used in a variety of reasoning and learning experiments.


Blass

AAAI Conferences

Moral reasoning is important to accurately model as AI systems become ever more integrated into our lives. Moral reasoning is rapid and unconscious; analogical reasoning, which can be unconscious, is a promising approach to model moral reasoning. This paper explores the use of analogical generalizations to improve moral reasoning. Analogical reasoning has already been used to successfully model moral reasoning in the MoralDM model, but it exhaustively matches across all known cases, which is computationally intractable and cognitively implausible for human-scale knowledge bases. We investigate the performance of an extension of MoralDM to use the MAC/FAC model of analogical retrieval over three conditions, across a set of highly confusable moral scenarios.