Knowledge representation and reasoning

About: Knowledge representation and reasoning is a research topic. Over the lifetime, 20078 publications have been published within this topic receiving 446310 citations. The topic is also known as: KR & KR².

Modular utility representation for decision-theoretic planning

Abstract: Specification of objectives constitutes a central issue in knowledge representation for planning. Decision-theoretic approaches require that representations of objectives possess a firm semantics in terms of utility functions, yet provide the flexible compositionality needed for practical preference modeling for planning systems. Modularity, or separability in specification, is the key representational feature enabling this flexibility. In the context of utility specification, modularity corresponds exactly to well-known independence concepts from multiattribute utility theory, and leads directly to approaches for composing separate preference specifications. Ultimately, we seek to use this utility-theoretic account to justify and improve existing mechanisms for specification of preference information, and to develop new representations exhibiting tractable specification and flexible composition of preference criteria.

Precedence knowledge in feature mating operation assembly planning

Abstract: The authors discuss the representation and acquisition of the precedence knowledge of an assembly, which plays an important role in the generation of assembly sequences and the planning of assembly. An efficient and complete symbolic representation has been developed to express the precedence knowledge clearly and precisely. This symbolic representation makes it possible to perform reasoning and manipulation of the precedence knowledge. Furthermore, the representation is complete in the sense that it can represent the assembly precedence knowledge as well as the disassembly precedence knowledge and these two forms of knowledge can be transformed from one to another. A geometric mating graph is developed to include all the necessary geometric and topological information for the precedence knowledge acquisition. Two algorithms are developed to obtain the precedence knowledge from the geometric mating graph systematically. The disassembly precedence knowledge thus obtained is equivalent to the assembly precedence knowledge and can be used to generate all the possible sets of assembly sequences. >

NSF workshop on Visual Information Management Systems

Abstract: One of the most important technologies needed across many traditional areas as well as emerging new frontiers of computing, is the management of visual information. For example, most of the Grand Challenge applications, under the High Performance Computing and Communication (HPCC) initiative, require management of large volumes of non-alphanumeric information, computations, communication, and visualization of results. Considering the growing need and interest in the organization and retrieval of visual and other non-alphanumeric information, and in order to stimulate academic projects in this area, a workshop on Visual Information Management Systems (VIMS) was sponsored by the National Science Foundation. This workshop was held in Redwood, CA, on February 24-25, 1992. The goal of the workshop was to identify major research areas that should be addressed by researchers for VIMS that would be useful in scientific, industrial, medical, environmental, educational, entertainment, and other applications.The major findings of the workshop were that VIMS require new techniques in all aspects of databases, computer vision, and knowledge representation and management; and that such techniques are best developed in the context of concrete, practical applications. VIMS will provide impetus and testbeds for many techniques being explored for the future database systems. Researchers from image processing and understanding, knowledge representation and knowledge based systems, and databases must work very closely to develop VIMS. Such systems should be developed in the context of applications that will be of immediate interest in industrial, medical, or scientific contexts. Without concrete applications and ambitious implementation projects, most of the important and difficult issues are likely to be ignored. Considering the interdisciplinary nature of the research in this area, a few major research projects in this area are essential for its growth. Increased emphasis on HPCC by many Federal agencies can help in the rapid development of VIMS technology. Similarly, by addressing some of the Grand Challenges, research interested in VIMS can understand critical issues and develop techniques to solve them, in a concrete and useful context.Parallel processing is essential for implementing VIMS. As is well known, the processing of images is one of the most computation-intensive tasks. For entering images in databases and for performing required operations at query time, an enormous volume of data must be processed. Parallel computing will be essential for implementing a VIMS that can insert images in reasonable time and provide fast response to user queries. The computational requirements of video databases are likely to be one of the most demanding. It is very likely that video databases will require research in highly parallel-pipelined architectures.In interdisciplinary research areas such as VIMS, most important and difficult problems usually fall through the cracks. The three most relevant areas for the development of VIMS are: databases, computer vision, and knowledge representation. Data compression, fault-tolerant real time access to image data through networks, and parallel processing issues should be addressed in the context of databases for VIMS. VIMS should not be considered as an application of the existing state of the art in any of these fields to manage and process images. Database researchers must understand the issues specific to managing and processing images and other forms of data by granting them the same status that has been given to alphanumeric information. Computer vision researchers should identify features required for interactive image understanding, rather than their discipline's current emphasis on automatic techniques, and develop techniques to compute features in interactive environments. Most knowledge representation research has been concerned with symbolic k knowledge. For VIMS and HPCC applications, techniques for representing symbolic and non-symbolic representations at the same level will be required. Reasoning approaches that can deal with such representations will be useful not only in VIMS, but in many other applications also. Finally, performance issues pose a significant challenge in all aspects of VIMS, from memory organization to information retrieval.

Question Answering as Global Reasoning Over Semantic Abstractions.

Abstract: We propose a novel method for exploiting the semantic structure of text to answer multiple-choice questions. The approach is especially suitable for domains that require reasoning over a diverse set of linguistic constructs but have limited training data. To address these challenges, we present the first system, to the best of our knowledge, that reasons over a wide range of semantic abstractions of the text, which are derived using off-the-shelf, general-purpose, pre-trained natural language modules such as semantic role labelers, coreference resolvers, and dependency parsers. Representing multiple abstractions as a family of graphs, we translate question answering (QA) into a search for an optimal subgraph that satisfies certain global and local properties. This formulation generalizes several prior structured QA systems. Our system, SEMANTICILP, demonstrates strong performance on two domains simultaneously. In particular, on a collection of challenging science QA datasets, it outperforms various state-of-the-art approaches, including neural models, broad coverage information retrieval, and specialized techniques using structured knowledge bases, by 2%-6%.

Using context to build semantics.

Abstract: Latent semantic analysis (LSA) is a model of knowledge representation for words. It works by applying dimension reduction to local co-occurrence data from a large collection of documents after performing singular value decomposition on it. When the reduction is applied, the system forms condensed representations for the words that incorporate higher order associations. The higher order associations are primarily responsible for any semantic similarity between words in LSA. In this article, a memory model is described that creates semantic representations for words that are similar in form to those created by LSA. However, instead of applying dimension reduction, the model builds the representations by using a retrieval mechanism from a well-known account of episodic memory.