File Information
File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/ackno/90/c90-3094_ackno.xml
Size: 17,434 bytes
Last Modified: 2025-10-06 13:51:41
<?xml version="1.0" standalone="yes"?> <Paper uid="C90-3094"> <Title>A Project Report on NP: an Assumption.based NL Plan Inference System that uses Feature Structures</Title> <Section position="2" start_page="0" end_page="0" type="ackno"> <SectionTitle> Abstract </SectionTitle> <Paragraph position="0"> This paper presents a project report on NP, a working Natural language Plan inference system that uses feature structures and is based on assumptions. Input to the system is in the form of feature structures, which can be taken directly from the output of a semantic parser. Plan actions are represented by feature-structure plan schemata with preconditions, hierarclfical decompositions, and effects. Output is in the form of a network of believed assertions represented in a knowledge base, and can be reported, used to answer generation-system queries, or drive side-effecting demons. The plan inference component is implemented using a feature-structure-based inference engine and models of plan recognition, prediction, and inference. The inference engine is implemented using a rewriting system for patteru-matctfing, and an Assumption-based Truth Maintenance System (ATMS) for conjunctions. The ArMS allows pre-instantiation of hypothetically known assertions and implications, which can significantly reduce processing time. The ATMS also permits simultaneous consideration of multiple possible inputs or multiple possible inferred plan outputs; these can be mutually conflicting or supportive.</Paragraph> <Paragraph position="1"> This capability will be important for disambiguation. The NP system is used to infer dialog- and domaln-level plans, among other types.</Paragraph> <Paragraph position="2"> Original contributions include: a plan inference system that works directly from feature structures; a plan inference system that uses an ATMS and plan schema actions with pre-conditions and effects to infer hierarchical and chained plans; and, an inference engine that works with multiple feature-structure assertions and rules.</Paragraph> <Paragraph position="3"> Project Goal. This project is aimed toward a dialog understanding system that can be used as part of an automatic interpreting telephone system.</Paragraph> <Paragraph position="4"> Interpretation will be performed by parsing, transferring, and generating utterances. Thus, dialog understanding will be used to recognize speech acts and illocutionary acts, resolve ellipses, and provide required missing information, among other tasks. The understanding system will use the output of the semantic parser, and provide information to the transfer module and generation system. Therefore, feature structures should be used as the basic data representation scheme. Dialog understanding requires a general-purpose plan inference engine that can work with dialog plans, domain plans, common-sense knowledge plans, and so forth. The system must also in the future be able to perform disambiguation of possible utterances.</Paragraph> <Paragraph position="5"> Background: Assumptions. The plans, intentions, beliefs, etc., of a human are menial concepts which cannot be perceived directly', they are unobservable\[Mye88\]. There is insufficient information to represent these concepts with certainty.</Paragraph> <Paragraph position="6"> Therefore, the system must be able to represent concepts in an uncertain manner, using assumptions.</Paragraph> <Paragraph position="7"> Communication between two people is inherently an assumption-based process. Since it is never cornpletely possible to directly know the concepts of another person, it is necessary in the course of a conversation to take a stance and rely on assumptions about the other person's thoughts \[Den87\]. Thus, in a dialog understanding system, there are at least two kinds of assumptions that must be represented: assumptions that the two speakers make, which must be modeled by the system, and assumptions that the system makes about the situation, 1 the two speakers and their plans, intentions, etc.</Paragraph> <Paragraph position="8"> Design. Plan inference and other knowledge-based reasoning tasks require that multiple conjunctive implications be matched against large sets of unordered assertions. The system will have a catalog of world knowledge, common-sense knowledge, and assertions which are believed by the speakers.</Paragraph> <Paragraph position="9"> These must be accessed non-sequentially and used for reasoning. In other words, language understanding should be done by using an &quot;expert system&quot; inference engine, Computer languages should be used according to their strengths and weaknesses. Feature-structure systems are strong in representing complex, incomplete, or underspecified information, and in performing unification. However, they are extremely inefficient at list processing and numeric calculations (e.g. for evidential reasoning), and don't represent multiple possible worlds. Lisp and other languages can fulfil these needs.</Paragraph> <Paragraph position="10"> One solution is to build a hybrid system. An inference engine was built which uses a feature-structure language fo r representation and pattern-matching tasks, while using an ATMS to perform conjunctive implications, represent assumptions, represent possible worlds, and maintain the truth of derived belief networks when nonmonotonic changes occur. The ATMS allows the system to represent, and reason with, all consistent possibilities at the same time-not just the current best choice.</Paragraph> <Paragraph position="11"> In particular, this permits multiple possible inferred plans to be output, and multiple possible observations to be input. This capability will become important for possible utterance disambiguation.</Paragraph> <Paragraph position="12"> The system interprets the results of the ATMS by using a five-valued uncertainty logic consisting of the uncertain belief values ACTUAL, POSSIBLE, HY-POTHETICAL, INCONSISTENT, or NULL. Each asser1 Currently, most dialog understanding systems start with the assumptions that the hearer and speaker always understand each other perfectly, that they automatically want to cooperate as much as possible, and that they have absolutely no other commitments outside of the conversation. Clearly some of these assumptions can occasionally be incorrect.</Paragraph> <Paragraph position="13"> \[\[action \[\[RELN tthort-antsuor-nogativo-sot-1\] ....... \[AGEN ?queat loner'\] \] \]. O--,~,c:{~ i \[procl \[ \[IILEI, H S-REQUEST\] ~ :':~&quot; t::i ........ i \[OBJE \[ \[REI.~ INFORMI F\] \[AGE// ?answerer\] , ,:>-a~ \[0BJE \[ \[RELN ?vorb\]</Paragraph> <Paragraph position="15"> tiou is assigned a vahw. Not(: that merely because an assc'rtion ix l>elieved (.rue in all l,:nc)wn consisl.('nt l)OS nil>h' worlds, this (lo<'s \]Jot mean that it is ACTL!AI, it might only bc VOSS~))LE. So<: \[Myc89a\] for (le(ails.</Paragraph> <Paragraph position="16"> Plan Schemata. The plan schemata are best explained by an actual cxamphx :~' See Figure 1. Plan ,~'ch('mata a.re lbrmed from possil)ly cyclic fi~al,ure strucl, ures and can incl.de varial>l,'s, co-iustanc(: variahh's aud rest varial)\[o,c A sch(.ma has a 1t;1111(% a s('ri('s of l>r(wolldili<)ns, a series o(&quot; (hx'omt>ositions, au<l a series of (~(l'ect.s.</Paragraph> <Paragraph position="17"> Plal~ Inf(q'ence. l)la)) infi'r<'m'e is l>e)'formc.d hy iml>h,m<')~(ing too<Ms for pla)) rec<>gnition, prediction, and iufi,rcnce on i.oI> <)film inferenc(&quot; engine (thus, it, also use~; fi'ature strm'tures as its maiu dai,a rcprescni,a(,ion). I{ecognii km ix based o. 'rcq.in:d on/ailment and use, a slrong mo(M i.\]mI, slates (rll~lt, the conju)wtion of f.h(' (h'coml><)si(.i<ms plus l>)'(~con(\]i fi<)ns iml>li('s r('cogni(.ion ,>f the ;wl.i<m (in a l>(>l.(onl: up fashion). If a w('alwr )*eco~;)fi(.io)) m('(.hod is de= sired, (lu' ust, r cat) Sl>ecil'y ,vu.\[lictcnc!j s<'ls (>f parl,ic-!liar l>r,'c<>,di(.ions aml d('COml><)siti<)))s t<> recog))ize Ih<' a<'l.i(>n. (l';.g., {t)re<:l, <1,'c2} is slllli<;i(')fl, for lh(' previous exa)nple.) ;~ If the enl.ailme))t is not )'cqutrrd, an assu))q>tion I,l.(.i the anl.e<'ed('nl.s d<)in fact imply the re<'ognized ac(.i<))) can l>(, hlc<)rpor;l.(.(><l in the con.j,)uc(.iou, .yi(,hling a c,>),liti(>,,~l r,'<x>gnil.i<m.</Paragraph> <Paragraph position="18"> The model fin&quot; I)lal~ I)rcdi<'ti<m re<l.ircs {,hat each are formed in Jalwtnese by rCl)e;tting |he vm'l). AII,h(mgh these are semantically and I)ragmalically wvll f.rmcd, (hey ,:am.n, I)c tran..,latcd liLcrally Iml must Iw rccogniv, ed and lrcumferred, sin.'c I';nglish f..vms sh,,r( ;mswcr.', by repealing Lhc auxiliary (&quot;No. I dos'l.&quot;). The kqv fcalmc is (h;~.l the w,b is rclw;dcd wit.h(>,~l, an ()l>je(:(., Ml.cr an i,)f(>rm-if (yes/no) qucstio,,. aThis is ~t re|il)('nl,ql(. ()f l'{)i,)h\]o<'k's ne(:css;uy ;u.I :;ulIi,:ien( (:,,,,,li(.i,,)m \[I<,.)SS\].</Paragraph> <Paragraph position="20"> Since the. hlferencc engine is ill turn based on an ATMS, the plan infereuce n'ethod also inherits (he capabilities of representing po,;sibh. (assulned) or actual asserlions. The representation carefully distinguishes between actual, possible, predicted, and hypol.hctical occurrences. M ul\[iple self-consistent possibilitics can be represented. This results in inforred plans also being possible or actual.</Paragraph> <Paragraph position="21"> Norm'ally, i( is i>ossible to implemen\[ other too<tel.'+ <,f recognilion, I>rediction, or inference, .sing the s)slx'm 'Fh<~ formal philosophical foum\]ali<>us of I>\]an inference are not w,'\]l und<.rstood 0&quot;.g, dir(.<q, imlire<'t, a)M int<'ra<'ting causes \[Pea88!; or (rigg,'r-ing vs. suppor(ing <'ousts), aml are being researchd, A <'a,'eful examination of the model sl>ocificatians shown hcrc reveals that the current systen~ infers plans having monotonic actions (although particular st.ares may be. retracted in a nomllonotonic fashion).</Paragraph> <Paragraph position="22"> N(mmonotonic extensions are being inw'stigated.</Paragraph> <Paragraph position="23"> Execution. There are three kinds of input to the systt'm. First, the plan <le,,dgn<:r specifies a series of plan s<:henlata. Next, workl knowh'dgc assc)'ticms and hyl>oMwtical utt,c:r;mces, in the form (ff f<'at.re stru<:t.res, a, re f<'d in(:o Lhc sy~t<.)J~ and l>r< * ins(an(iatd, l,ater, possiMe (can<lidatc) or actual ,tterances are i;~put into tim system and used for I>l;t,, inDrerwing= 'l'he resull.s <:onsisL of a n<:twork of iufer<mcas aml recognized plans, representc<t insi<h: the ATMS.</Paragraph> <Paragraph position="24"> 'l'his structured kn<>wle<tlre base can bc <luerie<l by the transfer or general, ion systems to provi<le iu-formatiol, 1,o resolve \]>rol>h'ms. AIterl,atively, sideeffe<:ting <h,m<ms attache<l to sele<'ted ATMS nodes can process and report plan infiq'cnce information auton<)t,l<),,sly, l)enlons are also used to set a.nd deleie processing flags in the ATMS network (e.g., for printing out the results of an inferred l>lan only olwe, ewm though t, he results are conl, immlly t, ru<').</Paragraph> <Paragraph position="25"> Techique. Inl)ul, plan schemata are fed to tlw I)lan--schema interpreter, which breaks i, hen~ Ul> and creat<~s an irish'action rule fo.r each \[>re<'on<lition or do<'Oml>osi(ion in (.h(' s<'lwma. An last.ruction rule c<))Isis(.s of a l>att.er)l (with va.riables) l.<) l><' r(x'(~g I)iz('d as an aul.eced('nt, l>lus insl.ru<'l.ions (wilh vm'i '2 429 ables) as the consequent. The instruction rules are fed to the nonmonotonic rewriting system \[Has89\] and used as rewriting rules.</Paragraph> <Paragraph position="26"> Preinstantiation assertions are input to the rewriting system. One assertion may match several instruction rule patterns. The recognized consequents consist of instructions with instantiated variables. These are fed to the instruction interpreter, which follows the instructions and instantiates hypothetical nodes and implications in the ATMS corresponding to instantiated components of the plan schema. The resulting action and effect assertions are fed back to the rewriting system as more hypothetical input, to instantiate whole networks bottom-up from single input facts. Since the ATMS uses &quot;uniquification&quot; and never reinstantiates existing assertions, this process eventually terminates.</Paragraph> <Paragraph position="27"> Run-time utterances are submitted to a feature-structure hash test which checks to see whether they have identically been instantiated in the ATMS before (e.g., hypothetically, etc.). If they have, no further pattern matching is required, and the corresponding node is asserted. Otherwise, the input utterance is submitted to the rewriting system for pattern matching and hypothetical instantiation propagation, as before. Following this, the corresponding node is asserted as possible or actual. Assertion triggers a fast spreading activation in the truth maintenance network (using bitvectors) which maintains all consistent &quot;possible worlds&quot;. A demon attached to a particular node fires when that node first becomes possible or actual. Demons can reset themselves by deleting support node flags.</Paragraph> <Paragraph position="28"> Current Status. NP version 2.0, reported here, has been finished and demonstrated. Currently instantiation is done in a bottom-up fashion. A graphic output program allows display of the ATMS network. Nodes can be moused for input or examination. A browsing editor needs to be improved to allow relevant parts of the network to be examined.</Paragraph> <Paragraph position="29"> The system currently works stand-alone and is ready to be integrated with a parser, a transfer system, and a language generator when they have been finished.</Paragraph> <Paragraph position="30"> Applications. The plan inference system understands ongoing task-oriented conversations between two people, on the subject of registering for a conference. The system serially processes the utterances, maintaining a representation of the currently believed concepts as the conversation progresses.</Paragraph> <Paragraph position="31"> Currently, input to the system is a corpus of five conversations (20 utterances each, on average) representing the expected feature-structure output of the ATR parsing system, as generated by the parsing researchers. To date, 53 plan schemata have been written, dealing with conversation opening and closing sequences, &quot;inform-if&quot; (yes/no) questions and answers, short answers to questions, ability utterances, inferred knowledge, wants and intentions, domain plans, idioms, and common-sense knowledge.</Paragraph> <Paragraph position="32"> The system understands portions of all five conversations. Research is ongoing in this area. In addition, a separate set of 46 plans duplicating the &quot;four-layer&quot; recognition model \[AI89\] in feature structures has been implemented and used to understand an entire surface-speech-act-level input conversation with 20 speech-acts (see Figure 2); 232 feature-structure assertions were instantiated.</Paragraph> <Paragraph position="33"> Future Work. A manual is being written. The instantiation method must be improved, l~epresenting multiple possible nonmonotonic plans is being researched. It will also be necessary to represent and infer plans containing conditional branches, for information-gathering plans in the sample dialogs 4.</Paragraph> <Paragraph position="34"> However, the main fimlre research deals with the design of a disambiguation system. Disambiguation is an evidential reasoning problem, and will probably require that a causal reasoning system be built and integrated with the multiple-world ATMS.</Paragraph> <Paragraph position="35"> Naturally, parallel research in illocutionary and perlocutionary force is also continuing. One of the first things that must be done is to build a module that uses possible plans to resolve zero pronouns.</Paragraph> <Paragraph position="36"> Lack of a fully resolved utterance is hindering plan recognition. Other work involves representing and reasoning with intentions found in a dialog.</Paragraph> <Paragraph position="37"> Coneluslon. A working plan inference system that uses feature structures has been d,.scribed. The system can represent and reasou with assumptions, multiple possible inputs, and multiple possible results. Preinstantiation of hypothetical assertions allows inference by fast spreading activation. The resulting system can be used to infer plans directly from the output of a sC/.naantic~, parser, and provide information directly to a transfer or generation system, as part of an automatic interpretation system.</Paragraph> <Paragraph position="38"> 4E.g., &quot;Do you have the announcement?&quot; &quot;No.&quot; &quot;Then I will give you the details:...&quot; 430 3</Paragraph> </Section> class="xml-element"></Paper>