This paper surveys research in developing computational models for integrating linguistic and visual information. It begins with a discussion of systems which have been actually implemented and continues with computationally motivated theories of human cognition. Since existing research spans several disciplines (e.g., natural language understanding, computer vision, knowledge representation), as well as several application areas, an important contribution of this paper is to categorize existing research based on inputs and objectives. Finally, some key issues related to integrating information from two such diverse sources are outlined and related to existing research. Throughout, the key issue addressed is the correspondence problem, namely how to associate visual events with words and vice versa. 1 Introduction Much has been said about the necessity of linking language and vision in order for a system to exhibit intelligent behaviour [Win73, Wal81]. A complete natural-language und...

Most natural language processing tasks require lexical semantic information. Automated acquisition of this information would thus increase the robustness and portability of NLP systems. This paper describes an acquisition method which makes use of fixed correspondences between derivational affixes and lexical semantic information. One advantage of this method, and of other methods that rely only on surface characteristics of language, is that the necessary input is currently available.

Abstract not found

Thematic knowledge is a basis of semantic interpretation. In this paper, we propose an acquisition method to acquire thematic knowledge by exploiting syntactic clues from training sentences. The syntactic clues, which may be easily collected by most existing syntactic processors, reduce the hypothesis space of the thematic roles. The ambiguities may be further resolved by the evidences either from a trainer or from a large corpus. A set of heurisL:cs based on linguistic constraints is employed to guide the ambiguity resolution process. When a train,-r is available, the system generates new sentences ,s'kose thematic vaiidities can be justified by the trainer. When a large corpus is available, the thematic validity may be justified by observing the sentences in the corpus. Using this way, a syntactic processor may become a thematic recognizer by simply deriving its thematic knowledge from its own syntactic knowledge.

In this paper we present a computational study of lexical acquisition. We attempt to characterize the lexical acquisition task faced by children by defining a simplified formal approximation of this task which we term the mapping problem. We then present anovel strategy for solving large instances of this mapping problem. This strategy is capable of learning the word-to-meaning mappings for as many as 10,000 words given corpora of 20,000 utterances. Such lexical acquisition is accomplished in a language independent fashion without any reference to the syntax of the language being learned.

Nondeterministic Lisp is a variant of Lisp with a nondeterministic choice operator. This manual describes an efficient implementation of nondeterministic Lisp called Screamer. Screamer is implemented as a fully portable macro package built on top of Common Lisp. Screamer functions inter-operate in the same environment as ordinary Lisp functions and a large subset of Common Lisp is available when writing Screamer functions. In addition to the nondeterministic choice operator, Screamer provides a forward checking constraint propagation facility as well. Together they make Screamer an efficient mechanism for building search programs. TOPIC AREAS: nondeterministic search, AI programming languages Caution! This product may drive you zonkers! From the box of Screaming Yellow Zonkers 1 Introduction 2 Nondeterministic Expressions, Functions and Contexts In order to provide the ability for backtracking, Screamer compiles nondeterministic functions differently than deterministic functi...

: This paper presents an overview of current research concerning knowledge extraction from technical texts. In particular, the use of empirical techniques during the identification and generation of a semantic representation is considered. A key step is the discovery of useful n-grams and correlations between clusters of these n-grams. keywords: knowledge representation, large text corpora, language understanding. 1. BACKGROUND The primary knowledge extraction and text retrieval conferences (MUC-4, 1992; TREC-1, 1993; TIPSTER, forthcoming) utilise domain-specific queries and templates to identify relevant concepts from within a corpus and extract applicable documents or information. The structures generated by the system discussed in this paper are similar to these domain-specific templates, they could be used for compact representation of information contained in documents for text retrieval purposes. The automatic generation of templates would be a significant development. The mo...

: A fundamental issue in natural language processing is the prerequisite of an enormous quantity of preprogrammed knowledge concerning both the language and the domain under examination. Manual acquisition of this knowledge is tedious and error prone. Development of an automated acquisition process would prove invaluable. This paper references and overviews a range of the systems that have been developed in the domain of machine learning and natural language processing. Each system is categorised into either a symbolic or connectionist paradigm, and has its own characteristics and limitations described. Key words: machine learning, natural language processing, cognitive modelling, knowledge acquisition, knowledge representation. page 2 1. INTRODUCTION 1.1 Artificial intelligence learning strategies The study of machine learning in fields such as Artificial Intelligence (AI), Psychology and Neurology has been intense, but has attained limited success. In AI a wide range of strategi...

Abstract not found

This paper describes two groups of experiments carried out with a Neural State Machine (NSM) [1], built using Weightless Artificial Neurons called General Neural Units (GNUs) [2], and simulated in near real-time using the interactive software simulator MAGNUS [3,4]. Visual input is provided to the system from real grey-scale video images, using the KITCHENWORLD environment. The experiments address the longstanding problem of how to ground linguistic symbols in order that artificial language processing systems, such as might be used with robots, can use language without needing a human intermediary to translate every input to the processor into its own closed symbol system.