The Workshop on Future Directions in NLP was held at Bolt Beranek and Newman, Inc. (BBN), in Cambridge, Massachusetts, from 29 November to 1 December 1989. The workshop was organized and hosted by Madeleine Bates and Ralph Weischedel of the BBN Speech and Natural Language Department and sponsored by BBN's Science Development Program. Thirty-six leading researchers and government representatives gathered to discuss the direction of the field of natural language processing (NLP) over the next 5 to 10 years. The intent of the symposium was "to make the conference and resulting volume an intellectual landmark for the field of NLP." This brief article summarizes the invited papers and strategic planning discussions of the workshop.

This article surveys the use of empirical, machinelearning methods for a particular natural language-understanding task--information extraction. The author presents a generic architecture for information-extraction systems and then surveys the learning algorithms that have been developed to address the problems of accuracy, portability, and knowledge acquisition for each component of the architecture. Author Eugene Charniak and coauthors Ng Hwee Tou and John Zelle, for example, describe techniques for part-of-speech tagging, parsing, and word-sense disambiguation. These techniques were created with no specific domain or high-level language-processing task in mind. In contrast, my article surveys the use of empirical methods for a particular natural language-understanding task that is inherently domain specific.

In recent years, there has been a flurry of research into empirical, corpus-based learning approaches to natural language processing (NLP). Most empirical NLP work to date has focused on relatively low-level language processing such as part-ofspeech tagging, text segmentation, and syntactic parsing. The success of these approaches has stimulated research in using empirical learning techniques in other facets of NLP, including semantic analysis--uncovering the meaning of an utterance. This article is an introduction to some of the emerging research in the application of corpusbased learning techniques to problems in semantic interpretation. In particular, we focus on two important problems in semantic interpretation, namely, word-sense disambiguation and semantic parsing.

Stanford Unzversity Stanford, CA 94805 F'OUNDED EARLY IN 1983, the Center for the Study of Language and Information [CSLI] at Stanford University grew out of a longstanding collaboration between scientists at research laboratories in the Palo Alto area and the facult,y and students of several Stanford University departments and out of a need for an institutional focus for this work on natural and computer languages. At present, CSLI has 17 senior members and about as many associate members, from SRI International, Xerox PARC, Fairchild, and the Departments of Computer Science, Linguistics, and Philosophy at Stanford. Since the Center's research will overlap with the work of other researchers around the world, an important goal of CSLI is to initiate a major outreach, whereby members of CSLI both inform themselves of work done elsewhere and share their own results with others. Questions about CSLI or Program SL should be addressed to Elizabeth Macken, Assistant Director, CSLI, Ven-tura Hall, Stanford University, Stanford, CA 94305. This collection of projects aims at developing scientific theories of natural-language use consonant with our basic perspective on language users as finite information processors.

In recent years, there has been a resurgence in research on empirical methods in natural language processing. These methods employ learning techniques to automatically extract linguistic knowledge from natural language corpora rather than require the system developer to manually encode the requisite knowledge. This article presents an introduction to the series of specialized articles on these topics and attempts to describe and explain the growing interest in using learning methods to aid the development of natural language processing systems. This special issue presents a machine-learning solution to the linguistic knowledge-acquisition problem: Rather than have a person explicitly provide the computer with information about a language, the computer teaches itself from online text resources. Since its inception, one of the primary goals of AI has been the development of computational methods for natural language understanding. Early research in machine translation illustrated the ...

BBN's project in knowledge representation for natural language understanding is developing techniques for computer assistance to a decision maker who is collecting information about and making choices in a complex situation. In particular, we are designing a system for natural language control of an intelligent graphics display. This system is intended for use in situation assessment and information management. Our work is concentrated in the development of fundamental techniques for knowledge representation and language understanding. Specifically, we are working on advanced parsing techniques, syntactic/semantic interaction, recognition of speaker intent, anaphora and deixis, fundamental knowledge representation techniques, and parallel algorithms and techniques for knowledge-based inference.

Several factors have led to the increase in interest in this field, which is heavily influenced by techniques from speech processing. One major factor is the recent availability of large online text collections. Another is a disillusionment with traditional AIbased approaches to parsing and natural language processing (NLP). Charniak is recognized as a distinguished contributor to what he calls traditional AI NLP, which is why it is all the more significant that in the Preface, when speaking of his recent transition to the statistical approach, he writes … few, if any, consider the traditional study of language from an artificial-intelligence point of view a "hot" area of research. A great deal of work is still done on specific NLP problems, from grammatical issues to stylistic considerations, but for me at least it is increasingly hard to believe that it will shed light on broader problems, since it has steadfastly refused to do so in the past.

It was motivated by two difficulties in scaling up existing generators. Current generators only accept input that are relatively poor in information, such as feature structures or lists of propositions; they are unable to deal with input rich in information, as one might expect from, for example, an expert system with a complete model of its domain or a natural language understander with good inference ability. Current generators also have a very restricted knowledge of language-- indeed, they succeed largely because they have few syntactic or lexical options available (McDonald 1987)-- and they are unable to cope with more knowledge because they deal with interactions among the various possible choices only as special cases. An utterance is simply the result of successive word choices. The treatment of syntax in connectionist and spreading activation systems is a well-known problem.

At the intersection of computational linguistics and artificial intelligence is where we find natural language processing. Very broadly, natural language processing (NLP) is a discipline which is interested in how human languages, and, to some extent, the humans who speak them, interact with technology. NLP is an interdisciplinary topic which has historically been the equal domain of artificial intelligence researchers and linguistics alike; perhaps obviously, those approaching the discipline from the linguistics side must get up to speed on technology, while those entering the discipline from the technology realm need to learn the linguistic concepts. It is this second group that this post aims to serve at an introductory level, as we take a no-nonsense approach to defining some key NLP terminology. While you certainly won't be a linguistic expert after reading this, we hope that you are better able to understand some of the NLP-related discourse, and gain perspective as to how to proceed with learning more on the topics herein.

It is important for intelligent systems, e.g. autonomous vehicles and robotics to anticipate the future in order to plan early and make decisions accordingly. Predicting the future scene parsing and motion dynamics helps the agents better understand the visual environment better as the former provides dense semantic segmentations, i.e. what objects will be present and where they will appear, while the latter provides dense motion information, i.e. how the objects move in the future. In this paper, we propose a novel model to predict the scene parsing and motion dynamics in unobserved future video frames simultaneously. Using history information (preceding frames and corresponding scene parsing results) as input, our model is able to predict the scene parsing and motion for arbitrary time steps ahead. More importantly, our model is superior compared to other methods that predict parsing and motion separately, as the complementary relationship between the two tasks are fully utilized in our model through joint learning. To our best knowledge, this is the first attempt in jointly predicting scene parsing and motion dynamics in the future frames. On the large-scale Cityscapes dataset, it is demonstrated that our model produces significantly better parsing and motion prediction results compared to well established baselines. In addition, we also show our model can be used to predict the steering angle of the vehicles, which further verifies the ability of our model to learn underlying latent parameters.