<?xml version="1.0" standalone="yes"?>
<Paper uid="P80-1003">
  <Title>ON THE EXISTENCE OF PRIMITIVE MEANING UNITS</Title>
  <Section position="4" start_page="0" end_page="0" type="metho">
    <SectionTitle>
2. WHAT IS A PRIMITIVE?
</SectionTitle>
    <Paragraph position="0"> All representation systems must have primitives of some sort, and we can see different types of primitives at different levels. Some primitives are purely structural and have little inherent associated semantics. That is, the primitives are at such a low level that there are no semantics pre-deflned for the primitives other than how they may combine. We call these primitives structural primitives. On the other hand, semantic primitives have both structural and semantic components.</Paragraph>
    <Paragraph position="1"> The structures are defined on a higher level and come with pre-attached procedures (their semantics) that indicate what they &amp;quot;mean,&amp;quot; that is, how they are to be meaningfully processed. What makes primitives semantic is this association of procedures with structures, since the procedures operating on the structures give them meaning. In a primitive-based scheme, we design both a set of structures and their semantics to describe a specific environment.</Paragraph>
    <Paragraph position="2"> There are two problems with pre-defining primitives.</Paragraph>
    <Paragraph position="3"> First, the choice of primitives may be structurally inadequate. That is, they may limit what can be represented. For example, if we have a set of rectilinear primitives, it is difficult to represent objects in a sphere world. The second problem may arise even if we have a structurally adequate set of primitives. I_n this case the primitives may be defined on too low a level to be useful. For example, we may define atoms as our primitives and specify how atoms interact as their semantics. Now we may adequately describe a rubber ball structurally, hut we will have great difficulty describing the action of a rolling ball. We would like a set of semantic primitives at a level both structurally and semantically appropriate to the world we are describing.</Paragraph>
  </Section>
  <Section position="5" start_page="0" end_page="13" type="metho">
    <SectionTitle>
3. INFERRING AN APPROPRIATE PRIMITIVE SET
</SectionTitle>
    <Paragraph position="0"> Schank \[1972\] has proposed a powerful primitive-based knowledge representation scheme called conceptual dependency. Several natural language understanding programs have been written that use conceptual dependency as their underlying method of knowledge representation. These programs are among the most successful at natural language understanding. Although Schank does not claim that his primitives constitute the only possible set, he does claim that some set of primitives is necessary in a general knowledge representation scheme.</Paragraph>
    <Paragraph position="1"> Our claim is that any advanced, sophisticated or rich memory is likely to be decomposable into primitives, since they seem to be a reasonable and efficient method for storing knowledge. However, this set of after-the-fact primitives need not be pre-defined or innate to a representation scheme; the primitives may be learned and therefore vary depending on early experiences.</Paragraph>
    <Paragraph position="2"> We really have two problems: inferring from early experiences a set of structural primitives at an appropriate descriptive level and learning the semantics to associate with these structural primitives. In this paper we shall only address the first problem. Even though we will not address the semantics attachment task, we will describe a method that yields the minimal structural units with which we will want to associate semantics. We feel that since the inferred structural primitives will be appropriate for describing a partitular environment, they will have appropriate semantics and that unlike pro-defined primitives, these learned primitives are guaranteed to be at the appropriate level for a given descriptive task. Identifying the structural primitives is the first step (probably a parallel step) in identifylng semantic primitives, which are composed of structural units and associated procedures that 81ve the structures meaning.</Paragraph>
    <Paragraph position="3"> This thesis developed while investigating learning strategies. Moran \[Salveter 1979\] is a program that learns frame-like structures that represent verb meanings. We chose a simple representative frame-like knowledge representation for Moran to learn. We chose a primitive-free scheme in order not to determine the level of detail at which the world must be described.</Paragraph>
    <Paragraph position="4"> As Moran learned, its knowledge base, the verb world, evolved from nothing to a rich interconnection of frame structures that represent various senses of different root verbs. When the verb world was &amp;quot;rich enough&amp;quot; (a heuristic decision), Moran detected substructures, which we call building blocks, that were frequently used in the representations of many verb senses across root verb boundaries. These building blocks can be used as after-the-fact primitives. The knowledge representation scheme thus evolves from a primitive-free state to a hybrid state. Importantly, the building blocks are at the level of description appropriate  Co how the world was described to Moran. Now Mor~ may reorganize the interconnected frames that make up the verb world with respect co the building blocks. This reorganizaclon renulcs in a uniform identification of the co--alleles and differences of the various meanings of different root: verbs. As l enrning continues the new knowledge incorporated into the verb world will also be scored, as ,-~ch as possible, with respect to the buildins blocks; when processing subsequent input, Moran first tries to use a on~inatlon of the building blocks to represent the meaning of each new situation iC</Paragraph>
    <Paragraph position="6"> A sac of building blocks, once inferred, need noc be fixed forever; the search for more building blocks may continue as the knowledge base becomes richer. A different, &amp;quot;better,&amp;quot; set of building blocks may be inferred later from the richer knowledge and all knowledge reorganized with respect to them. If we can assume that initial inputs are representaClve of future inputs, subsequent processing will approach that of primitive-based systems.</Paragraph>
  </Section>
  <Section position="6" start_page="13" end_page="14" type="metho">
    <SectionTitle>
4. AN OVERVIEW OF MORAN
</SectionTitle>
    <Paragraph position="0"> Moran is able to &amp;quot;view&amp;quot; a world that is a room; the room Contains people and objects, Moran has pre-defined knowledge of the contents of the room. For exan~le, it knows chac lamps, cables and chairs are all types of furniture, Figaro is a male, Ristin is a female, Eistin and Figaro are human. As input to a learning crlal, Moran is presented with:  i) a snapshot of the room Just before an action oct%tEn 2) a snapshot of tbe room Just after the action is completed end 3) a parsed sentence thac describes the action thac  occured in the two-snapshot sequence.</Paragraph>
    <Paragraph position="1"> The learning task is to associate a frame-like structure, called a Conceptual Meaning Structure (CMS), with each root verb it enco,mcers. A CMS is a directed acyclic graph that represents the types of entities chat participate in an action and the changes the entities undergo during the action.</Paragraph>
    <Paragraph position="2"> The ~s are organized so thac the similarities among various senses of a given root verb are expllcicly represented b 7 sharing nodes in a graph. A CMS is organized into two par~s: an ar~,-~-cs graph and an effects graph. The arguments graph stores cases and case slot restrictions, the effects graph stores a description of what happens co the entities described in the arg,,m~,~Cs graph when an action &amp;quot;takes place.&amp;quot; A sin~llfled example of a possible ~S for the verb &amp;quot;throw&amp;quot; is shown in Figure i. Sense i, composed of argument and effect nodes labelled A, W and X can represent '~kr 7 throws the ball.&amp;quot; Ic show thac during sense 1 of the actlan &amp;quot;throw,&amp;quot; a human agent remains at a location while a physical object changes location from where the Agent is to another location. The Agent changes from being in a stare of physical contact with the Object co not being in physical contact with ic. Sense 2 is composed of nodes labelled A, B, W and Y; It might represent &amp;quot;Figaro throws the ball co E-Istin.&amp;quot; Sense 3, composed of nodes labelled A, B, C, W, X and Z, could represent &amp;quot;Sharon threw the terminal at Raphael.&amp;quot; Mor~- infers a CMS for each root verb it encotmters.</Paragraph>
    <Paragraph position="3"> Although similarlt~'es among different senses of the same root verb are recognized, similarities are noC recognized across C~S boundaries; true synonyms might have id~-tlcal graphs, but Moran would have no knowledge  of the similarity. Similarities among verbs that are close in meaning, but not synonyms, are not represented; the fact that &amp;quot;move&amp;quot; and &amp;quot;throw&amp;quot; are related is not obvious to Moran.</Paragraph>
  </Section>
class="xml-element"></Paper>