Common sense has always been of interest in AI, but has rarely taken center stage. Despite its mention in one of John McCarthy's earliest papers and years of work by dedicated researchers, arguably no AI system with a serious amount of general common sense has ever emerged. Why is that? What's missing? Examples of AI systems' failures of common sense abound, and they point to AI's frequent focus on expertise as the cause. Those attempting to break the brittleness barrier, even in the context of modern deep learning, have tended to invest their energy in large numbers of small bits of commonsense knowledge. But all the commonsense knowledge fragments in the world don't add up to a system that actually demonstrates common sense in a human-like way. We advocate examining common sense from a broader perspective than in the past. Common sense is more complex than it has been taken to be and is worthy of its own scientific exploration.

It is a truism that as the Web grows in size and scope, it becomes harder to find what we want, to identify like-minded people and communities, to find the best ads to offer, and to have applications work together smoothly. Services don't interoperate; queries yield long lists of results, most of which seem to miss the point. If the Web were a person, we would expect richer and more successful interactions with it - interactions that were, quite literally, more meaningful. That's because in human discourse, it is shared meaning that gives us real communication. Yet with the current Web, meaning cannot be found.Much recent work has aspired to change this, both for human-machine interchange and machine-machine synchronization. Certainly the "semantic web" looks to add meaning to our current simplistic matching of mere strings of characters against mere "bags" of words. But can we legislate meaning from on high? Isn't meaning organic and determined by use, a moving and context-dependent target? But if meaning is an evolving organic soup, how are humans able to get anything done with one another? Don't we love to "define our terms"? But then again, is real definition even possible?These questions have daunted philosophers for years, and we probably won't solve them here. But we'll try to understand what's at the root of our own current religious debate: should meaning on the Web be evolutionary, driven organically through the bottom-up human assignment of tags? Or does it need to be carefully crafted and managed by a higher authority, using structured representations with defined semantics? Without picket signs or violence (we hope), our panelists will explore the two extreme ends of the spectrum - and several points in between.

Knowledge-base management systems (KBMS) based on description logics are being used in a variety of situations where access is needed to large amounts of data stored in existing relational databases. We present the architecture and algorithms of a system that converts most of the inferences made by the KBMS into a collection of SQL queries, thereby relying on the optimization facilities of existing DBMS to gain efficiency, while maintaining an object-centered view of the world with a substantive semantics and significantly different reasoning facilities than those provided by Relational DBMS and their deductive extensions. We address a number of optimization issues that arise in the translation process due to the fact that SQL queries with different syntax (but identical semantics) are not treated uniformly by current database management systems.

Corporate databases increasingly are being viewed as potentially rich sources of new and valuable knowledge. Various approaches to “discovering” or “mining” such knowledge have been proposed. Here we identify an important and previously ignored discovery task, which we call data archaeology. Data archaeology is a skilled human task, in which the knowledge sought depends on the goals of the analyst, cannot be specified in advance, and emerges only through an iterative process of data segmentation and analysis. We describe a system that supports the data archaeologist with a natural, object-oriented representation of an application domain, a powerful query language and database translation routines, and an easy-to-use and flexible user interface that supports interactive exploration. A formal knowledge representation system provides the core technology that facilitates database integration, querying, and the reuse of queries and query results.

CLASSIC is a data model that encourages the description of objects not only in terms of their relations to other known objects, but in terms of a level of intensional structure as well. The CLASSIC language of structured descriptions permits i) partial descriptions of individuals, under an 'open world' assumption, ii) answers to queries either as extensional lists of values or as descriptions that necessarily hold of all possible answers, and iii) an easily extensible schema, which can be accessed uniformly with the data. One of the strengths of the approach is that the same language plays multiple roles in the processes of defining and populating the DB, as well as querying and answering.

Knowledge Representation (KR) systems provide support for Artificial Intelligence systems that reason about relationships between objects in their domains of expertise. Because of their support for inference, KR systems appear to have potential to enrich the kind of retrievals that IR systems might make. Ironically, however, the most useful KR systems are limited to reasoning based on a rigid notion of validity, and thus are awkward to use when relevant but inexact retrievals are desired. We have been exploring the potential of a “connectionist” model—the Boltzmann Machine—to overcome this limitation. We report on a number of experiments in which we use a connectionist simulator to support similarity-based reasoning in a frame representation. We draw some tentative, mixed conclusions on the potential for a union of KR, IR, and connectionism.

A fundamental computational limit on automated reasoning and its effect on knowledge representation is examined. Basically, the problem is that it can be more difficult to reason correctly with one representational language than with another and, moreover, that this difficulty increases dramatically as the expressive power of the language increases. This leads to a tradeoff between the expressiveness of a representational language and its computational tractability. Here we show that this tradeoff can be seen to underlie the differences among a number of existing representational formalisms, in addition to motivating many of the current research issues in knowledge representation.

This paper describes the system and points out some of the benefits of using KLONE for representation in natural language processing