The FINITE STRING Newsletter
Site Report the esprit project loki
from several natural language technology base contracts with darpa's strategic computing program overview
Lt. Col. Robert L. Simpson Information Science Technology Office Defense Advanced Research Projects Agency
The overall objective of the Strategic Computing (SC) Program of the Defense Advanced Research Projects Agency (DARPA) is to develop and demonstrate a new generation of machine intelligence technology that can form the basis for more capable military systems in the future and also maintain a position of world leadership for the US in computer technology. Begun in 1983, SC represents a focused research strategy for accelerating the evolution of new technology and its rapid prototyping in realistic military contexts. The more specific top level goals supporting this single broad objective are to produce technology that will:
1. enable the operation of military systems under critical constraints such as time, information overload, etc.,
2. enable the management of forces/resources under constraints of information overload, geographic distribution, cost of operation, etc., and
3. facilitate the design, manufacture, and maintenance of defense systems within time, performance, quality, reliability, and cost constraints.
Even though capabilities for man-machine interaction will ultimately form an important component of systems in all of these areas, the second of those goals has been selected as the initial area to include emphasis on decision-making aids, including natural language processing. Subgoals of these top level goals include:
1. To strengthen/develop areas of science and technology that enable the building of computer systems needed to attain the top level goals. The technologies identified are: • Artificial Intelligence,
Site Reports
• Software development and machine architectures,
• Micro-electronics, and related infrastructure.
2. To build demonstration systems in specific military areas that:
• Provide focus for technology development,
• Provide means for exercising technology in real environments,
• Facilitate manpower training,
• Facilitate development of industrial capability, and
• Facilitate technology transfer to the military.
There are four very ambitious demonstration prototypes being developed within the SC Program. They are:
1. the Pilot's Associate, which will aid the pilot in route planning, aerial target prioritization, evasion of missile threats, and aircraft emergency safety procedures during flight;
2. the Autonomous Land Vehicle (ALV), which integrates in a major robotic test bed the technologies for dynamic image understanding, knowledge-based route planning with replanning during execution, hosted on new advanced parallel architectures;
3. two battle management projects: one for the Army, which is just getting started, called the AirLand Battle Management program (ALBM), which will use knowledge-based systems technology to assist in the generation and evaluation of tactical options and plans at the Corps level;
4. and the other more established program for the Navy, the Fleet Command Center Battle Management Program (FCCBMP) at Pearl Harbor. The FCCBMP is employing knowledge-based systems and natural language technology in an evolutionary test bed situated in an operational command center to demonstrate and evaluate intelligent decision aids that can assist in the evaluation of fleet readiness and explore alternatives during contingencies. It is within this context that the natural language contractors are currently demonstrating the potential of natural language technology.
Competitive awards were made to seven contractors in 1984. Four (BBN Laboratories, USC/Information Sciences Institute, the University of Pennsylvania, and the University of Massachusetts) are involved in research and development in natural language interfaces; three (New York University, Systems Development Corporation, and SRI International) are involved in research and development in text processing.
The work focuses on producing and demonstrating two "new generation systems": one for natural language interfaces and another for processing text in free form from military messages. BBN Laboratories serves as the integration contractor in natural language interfaces; New York University serves as the integration contractor in message processing. The remaining contractors are working on various component technologies, directly or indirectly contributing to the two new generation systems.
BBN LABORATORIES
Research and Development in Natural Language Processing in the Strategic Computing Program BBN Laboratories, Inc. Cambridge, MA 02238
Staff: Ralph Weischedel (Principal Investigator),
Remko Scha, Edward Walker, Damaris Ayuso, Andrew Haas, Erhard Hinrichs, Robert Ingria, Lance Ramshaw, Varda Shaked, David Stallard
BBN's responsibility is to conduct research and development in natural language interface technology. This responsibility has three aspects:
• to demonstrate state-of-the-art technology in a Strategic Computing application, collecting data regarding the effectiveness of the demonstrated heuristics,
• to conduct research in natural language interface technology, as itemized in the description of JANUS later in this note, and
• to integrate technology from other natural language interface contractors, including USC/Information Sciences Institute, the University of Pennsylvania, and the University of Massachusetts.
Of the three initial applications described in the overview, the Fleet Command Center Battle Management Program (FCCBMP) has been the application providing the domain in which our work is being carried out. The FCCBMP encompasses the development of expert system capabilities at the Pacific Fleet Command Center in Hawaii, and the development of an integrated natural language interface to these new capabilities as well as to the existing data bases and graphic display facilities. BBN is developing a series of increasingly sophisticated natural language understanding systems which will serve as an integrated interface to several facilities at the Pacific Fleet Command Center: the Integrated Data Base (IDB), which contains information about ships, their readiness states, their capabilities, etc.; the Operations Support Group Prototype (OSGP), a graphics system, which can display locations and itineraries of ships on maps; and the Force Requirements Expert System (FRESH), which is being built by Texas Instruments.
The target users for this application are naval officers involved in decision making at the Pacific Fleet Command Center; these are executives whose effort is better spent on navy problems and decision making than on the details of which software system offers a given information capability, how a problem should be divided to make use of the various systems, or how to synthesize the results from several sources into the desired answer. Currently they do not access the data base or OSGP application programs themselves; instead, on a round-the-clock basis, two operators act as intermediaries between the Navy staff and the computers. The utility of
Site Reports
a natural language interface in such an environment is clear.
The starting point for development of the natural language interface system at the Pacific Fleet Command Center was the IRUS system, which has been under development at BBN for a number of years. A new version of this system, IRUS-86, has been installed in the FCCBMP testbed area at the Pacific Fleet Command Center for demonstration. Further basic research on the problems of natural language interfacing is continuing, and the results of this and future research will be incorporated into a next generation natural language interface system called JANUS to be delivered to the Pacific Fleet Command Center at a later date. JANUS will share most of its domain-dependent data with IRUS-86, and it will share other modules as well; IRUS-86 will therefore be able to evolve gradually into the final version of JANUS.
The architecture of IRUS (Bates and Bobrow 1983) is a cascade consisting of a sequence of translation modules:
• An ATN parser, which produces a syntactic tree.
• A semantic interpreter, which produces a formula of the meaning representation language MRL.
• A postprocessor for resolving anaphora and ellipsis.
• A translation module, which produces a formula of the relational database language ERL (Extended Relational Language).
• A translation module, which produces a sequence of commands for the underlying database access system. Now installed at the Pacific Fleet Command Center,
IRUS-86 is a version of IRUS that has been extended in several ways. Two of these extensions are especially worth mentioning:
. IRUS-86 uses the NIKL system (Moser 1983) to represent its domain model, i.e., the relationships between the predicates and relations of the meaning representation language MRL. The NIKL domain model supports the system's treatment of semantic anomaly, anaphora, and nominal compounds.
• IRUS-86 contains a new module that exploits this NIKL domain model to simplify MRL expressions; this makes it possible to translate complex MRL-expressions into ERL constraints, thus allowing for significant divergences between the input English and the structure of the underlying data base (Stallard 1986).
In addition to accessing the NIKL domain model, the parser, semantic interpreter, and MRL-to-ERL translator access other knowledge sources that contain domain-dependent information:
• the lexicon,
• the semantic interpretation rules for individual concepts,
• the MRL-to-ERL mapping rules for individual MRL constants, which introduce the details of underlying system structure, such as file and field names.
To port IRUS to the Navy domain, the relevant domain-dependent data had to be supplied to the system.
This task is being accomplished by personnel at the Naval Ocean Systems Center (NOSC). In August 1985, BBN provided NOSC with an initial prototype system containing small example sets of lexical entries, semantic interpretation rules, and MRL-to-ERL rules; using acquisition tools provided by BBN, NOSC personnel have been entering the rest of the data.
IRUS-86 was delivered to the FRESH developers at Texas Instruments in January 1986, was installed in a testbed area of the Pacific Fleet Command Center in April 1986, and will be demonstrated in June 1986. Currently, the lexicon and the domain-dependent rules of the system only cover a relatively small part of the OSGP capabilities and the files and attributes of the Integrated Data Base. Once enough data have been entered so that the system covers a sufficiently large part of the data base, it will be tried out in actual use by Navy personnel. This will enable us to gather data about the way the system performs in a real environment, and to fine-tune the system in various respects. For instance, IRUS-86 makes use of shallow heuristic methods to address some aspects of natural language understanding such as anaphora and ellipsis for which general solutions are still research issues. The FCCBMP application provides a test bed in which such heuristic methods can be evaluated, and enhancements to them developed and tested, as part of the evolutionary technological growth intended to continue throughout the Natural Language Technology effort of the Strategic Computing Program.
The IRUS-86 system excels by its clean, modular structure, its broad syntactic/semantic coverage, its sophisticated domain model, and its systematic treatment of discrepancies between the English lexicon and the database structure. We thus expect that it will demonstrate considerable utility as an interface component in the FCCBMP application. Nevertheless, IRUS-86 shares with other current systems several limitations that should be overcome if natural language interfaces are to become truly "natural". In developing JANUS, the successor of IRUS-86, we shall attempt to overcome some of those limitations. The areas of increased functionality we are considering are: semantics and knowledge representation, ill-formedness, discourse, cooperativeness, multiple underlying systems, and knowledge acquisition.
IRUS-86, like most other current systems, represents sentence meanings as formulas of a logical language that is a slight extension of first-order logic. As a consequence, many important phenomena in English have no equivalent in the meaning representation language, and cannot be dealt with correctly, e.g., modalities, proposi-tional attitudes, generics, collective quantification, and context-dependence. Thus, one foregoes one of the most important potential assets of a natural language interface:
Site Reports
the capacity of expressing complex semantic structures in a succinct and comfortable way.
In JANUS, therefore, we will adopt a new meaning representation language that combines features from PHLIQAl's enriched lambda-calculus (Scha 1976) with ideas underlying Montague's Intensional Logic (Montague 1970), and possibly a distributed quote-operator (Haas 1986). It will have sufficient expressive power to incorporate a version of Carlson's treatment of generics (Carlson 1979), a version of Scha's treatment of quantification (Scha 1981), Montague's treatment of modality, and various possible approaches to proposi-tional attitudes and context-dependence.
In adopting a higher order logic as proposed, one confronts problems of formula simplification and the need to apply meaning postulates to reduce the semantic representation of an input sentence to an expression appropriate to the underlying system, e.g., a relational algebra expression in the case that the underlying system is a data base. To do this, we will investigate the limited inference mechanisms of KL-TWO (Moser 1983, Vilain 1985) , following up on our previous work (Stallard 1986) . The advantage of these inference mechanisms is their tractability; discovering their power and limitations in this complex problem domain should be an interesting result.
The meaning of a sentence depends in many ways on the context set up by the preceding discourse. IRUS and other systems, however, currently ignore most oi these dependencies, and employ a rather shallow model of discourse structure. To allow the user to exploit the full expressive potential of a natural language interaction, the system must track topics, reference times, possible antecedents for anaphora, etc.; it must be able to recognize the constituent units of a discourse and the subordination or coordination relations obtaining between them. A substantial amount of work has been done already on several of these issues, much of it by BBN researchers (Sidner 1985, Hinrichs 1981, Polanyi 1984, Grosz and Sidner 1986). Research in this area continues under a separate DARPA-funded contract. We expect to be able to integrate some of the results of that research in the JANUS system.
A natural interface system should be forgiving of a user's deviations from its expectations, be they misspellings, typographical errors, unknown words, poor syntax, incorrect presuppositions, fragmentary forms, or violated selection restrictions. Empirical studies show that as much as 25% of the input to database query systems is ill-formed.
IRUS currentlyv handles some classes of ill-formedness by using a combination of shallow heuristics and user interaction. It can correct for typographical misspellings, for omitted determiners or prepositions, and for some ungrammaticalities, like determiner-noun and subject-verb disagreement. The JANUS system will employ a more general approach to ill-formedness that will handle a larger class of ungrammatical constructions, a larger class of word selection problems, and that will also explore correcting several types of semantic ill-formedness.
These capabilities have major implications for the control of the understanding process, since considering such possibilities can exponentially expand the search space. Maintaining control will require care in integrating the ill-formedness capability into the rest of the system, and also in making maximal use of the guidance that can be derived from a model of the discourse and user's goals to constrain the search.
A truly helpful system should react not to the literal meaning of a sentence but to its perceived intent. If in the context of a given application it is possible to characterize the goals that a user may be expected to be pursuing through his interaction with the system, the system should try to infer from the user-input what the underlying goal could be. A system can do this by accessing a goal-subgoal hierarchy that links the speech acts expressed by individual utterances to the global goals that the user may have. This strategy has been applied successfully to rather small domains (Allen 1983, Sidner 1985). We wish to investigate whether it carries over to the FCCBMP applications.
The way in which IRUS-86 decides whether an input sentence translates into an IDB query or an OSGP command may be refined. There is a need for work on what kind of knowledge would be necessary to interface smoothly and intelligently to multiple underlying systems. A reasoning component is needed that can determine which underlying system or systems can best fulfill a user's request. Such a reasoning component would have to combine a model of the capabilities of the underlying systems with a model of the user goals and current intentions in the discourse context in order to choose the correct system(s). Such a model would also be useful for providing supporting information to the user.
Further research is also called for to expand the power of the knowledge acquisition tools used in adding to the lexicon, the set of case frame rules, the model of domain predicates, and the set of transformation rules between the Meaning Representation Language and the languages of the underlying systems. The acquisition tools available in HUJS, unlike those in some other systems, are not tied to the specific fields and relations in the underlying data base. The acquisition tools should work on the higher level of the domain model, since that provides a more general and transportable result. The knowledge acquisi-
Site Reports
tion facilities for JANUS will also need to be redesigned to support and to make maximal use of the power of the new meaning representation language based on inten-sional logic.
The new functionalities we described in the previous section, and the techniques we intend to use to achieve them, raise an issue that has important consequences for the design of JANUS: we will be faced with an explosion in the number of interpretations that the system will have to process; every sentence will be manifold ambiguous. One source of this phenomenon is the improvement of the semantic coverage and the broadening of the discourse context. Distinctions and ambiguities which so far were ignored will be dealt with: for instance, different interpretation and scopes of quantifiers will be considered, and different antecedents for pronouns. Even more serious is the processing of ill-formed sentences, which may require that some constraint be relaxed, while the only way to find out which one is to try all partial interpretations to see which one can be extended to a complete interpretation after relaxing one or more constraints.
To cut down on the processing of spurious interpretations, it is very important that interpretations of sentences and their constituents be tested for plausibility at an early stage. Different techniques must probably be used in conjunction:
• Simplification transformations may show that an interpretation is absurd, by reducing it to true or false or the empty set.
• The discourse context and the model of the user's goals impose constraints on expected sentences.
Since some of the techniques we intend to use to fight the ambiguity explosion are themselves rather computation-intensive, it is clearly unavoidable that the improved system functionality we aim for will lead to a considerable increase in the amount of processing required. To avoid a serious decrease of the new system's response times, we will therefore move it to a suitable parallel machine such as BBN's Butterfly or Monarch, running a parallel COMMON LISP. This in itself has rather serious consequences for the software design. It means that from the outset we will keep parallelizability of the software in mind.
We have begun to address this issue in the area of syntax. A new declarative grammar is being written, which will ultimately have a coverage of English larger than the current RUS grammar. The grammar is written in a side-effect-free formalism (a context-free grammar with variables) so that we may explore different parsing algorithms that are easily parallelizable. The first such algorithm was implemented in May 1986 on BBN's Butterfly.
We should not expect that JANUS will always be able to assess correctly which interpretation of a sentence is the intended one. In light of such situations, it is very important that the system can give a paraphrase of the input to the user, which shows the system's interpretation. This may be done either explicitly or as part of the answer. To be able to develop such capabilities, work on Natural Language Generation is needed. At USC/ISI a project directed by William Mann and Norman Sondheimer is underway to develop the generation system PENMAN, using the NIGEL systemic grammar. PENMAN will be integrated to become the generation component of JANUS. PENMAN itself consists of several subcomponents. Some of these, specifically the "text planning" component, will be developed through joint work between USC/ISI and David McDonald at the University of Massachusetts, based on the latter's experience with the MUMBLE system.
Under the direction of Aravind Joshi and Bonnie Webber of the University of Pennsylvania, several focussed studies have been carried out at UPenn to investigate various aspects of cooperative system behaviour and clarification interactions. (For more detail, see their report below.) As part of the Strategic Computing Natural Language effort, UPenn will eventually develop this into a module that can be integrated into JANUS to further enhance its capabilities.
LEXICON | WD LOOKUP | |||
Call for Papers
concentrated on the role of prior discourse, and has developed a procedure that handles a wide variety of noun phrase types, including pronouns and missing noun phrases, using a focusing algorithm based on surface syntactic structure (Dahl, submitted for publication). NYU, as part of its work on the domain model, has developed a procedure that can identify a component in the model from any of the noun phrases that can name that component (Ksiezyk and Grishman, submitted for publication). After further development, these procedures will be integrated into a comprehensive noun phrase semantic analyzer.
SDC has started to develop a module to process time information. Sources of time information include verb tense, adverbial time expressions, prepositional phrases, co-ordinate and subordinate conjunctions. These are all mapped into a small set of predicates expressing a partial time ordering among the states and events in the message.
The domain model captures the detailed information about the general class of equipment, and about the specific pieces of equipment involved in the messages; this information is needed in order to fully understand the messages. The model integrates part/whole information, type/instance links, and functional information about the various components (Ksiezyk and Grishman, submitted for publication).
The knowledge base performs several functions:
• It provides the domain-specific constraints needed for the semantics to select the correct arguments for a predicate, so that modifiers are correctly attached to noun phrases.
• It enables noun phrase semantics to identify the correct referent for a phrase.
• It provides the prototype information structures which are instantiated in order to record the information in a particular message.
• It provides the information on equipment structure and
function used by the discourse rules in establishing probable causal links between the sentences. And finally, associated with the components in the knowledge base are procedures for graphically displaying the status of the equipment as the message is interpreted. These functions are performed by a large network of frames implemented using the Symbolics Zetalisp flavors system.
4.7. DISCOURSE ANALYSIS (NYU)
The semantic analyzer generates separate semantic representations for the individual sentences of the message. For many applications it is important to establish the (normally implicit) intersentential relationships between the sentences. This is performed by a set of inference rules that (using the domain model) identify plausible causal and enabling relationships among the sentences. These relationships, once established, can serve to resolve some semantic ambiguities. They can also supplement the time information extracted during semantic analysis and thus clarify temporal relations among the sentences.
The diagnostic procedures are intended to localize the cause of failure of the analysis and provide meaningful feedback when some domain-specific constraint has been violated. We are initially concentrating on violations of local (selectional) constraints, and have built a small component for diagnosing such violations and suggesting acceptable sentence forms; later work will study more global discourse constraints.
Dahl, Deborah A. (submitted for publication) Focusing and Reference Resolution in PUNDIT.
Grishman, Ralph; Ksiezyk, Tomasz, and Nhan, Ngo Thanh (submitted for publication) Model-based Analysis of Messages about Equipment.
Hirschman, Lynette and Puder, Karl 1986 Restriction Grammar: A PROLOG Implementation. In Warren, D.H.D. and Van Caneghem, M., Eds., Logic Programming and its Applications. Ablex Publishing Company, Norwood, New Jersey: 244-261.
Hirschman, Lynette (in press) "Conjunction in Meta-Restriction Grammar." Journal of Logic Programming.
Ksiezyk, Tomasz and Grishman, Ralph (submitted for publication) An Equipment Model and its Role in the Interpretation of Nominal Compounds.
Palmer, Martha S. 1985 Driving Semantics for a Limited Domain.
Ph.D. thesis. University of Edinburgh. Palmer, Martha; Dahl, Deborah; Schiffman, Rebecca; Hirschman, Lynette; Linebarger, Marcia; and Dowding, John 1986 Recovering Implicit Information. To appear in Proceedings of the 24th Annual
Meeting of the Association for Computational Linguistics. Sager, Naomi and Grishman, Ralph 1975 The Restriction Language for Computer Grammars of Natural Language. Communications of Computer Grammar of English and its Applications. Addison-Wesley, Reading, Massachusetts.
Call for Papers escol 86
10-12 October 1986, University of Pittsburgh and Carnegie-Mellon University
The 1986 Eastern States Conference on Linguistics will include demonstrations of natural language processing software. The invited speakers are Charles Fillmore and Lily Wong Fillmore from the University of California at Berkeley, Martin Kay from the Xerox Palo Alto Research Center, and George Miller from Princeton University.
Original, unpublished papers on any topic of general linguistic interest are invited for the general sessions. For the special session, Linguistics at Work, we invite papers on applied linguistics, especially in the areas of language teaching and computational linguistics.