<?xml version="1.0" standalone="yes"?>
<Paper uid="C82-1006">
  <Title>o.&lt; Verb &gt; &lt; Location Obj.&gt; &lt; Indirect Obj.&gt; &lt; Dimension Obj.&gt; I~</Title>
  <Section position="3" start_page="0" end_page="0" type="metho">
    <SectionTitle>
I. PARSER QUALIFICATIONS
</SectionTitle>
    <Paragraph position="0"> The parser to be described performs its analysis starting from an intrinsically unreliable input that is the result of an isolated word speech recognizer. The lack of certainty on the single items of the input sentence is one of the main problems in such a vocal parser. The representation of each uttered word, following the recognition stage, is, in fact, an ordered list of possible interpretations with associated dissimilarity measures. As a consequence, it is possible to have doubts not only about every single word of the sentence, but also on complete sentence parts. Moreover, irrecoverable recognition errors may require the capability of parsing incomplete sentences.</Paragraph>
    <Paragraph position="2"/>
  </Section>
  <Section position="4" start_page="0" end_page="37" type="metho">
    <SectionTitle>
EXAMPLE OF A TYPICAL INPUT OF THE PARSER
</SectionTitle>
    <Paragraph position="0"> Fig. 1 shows an example of a typical parsing input where each input word is replaced by a complete list of possible alternatives with associated distance score. It  38 L. BORGHESI and C. FAVARETO is interesting to notice that not only the sentence that was actually spoken &amp;quot;TO-</Paragraph>
  </Section>
  <Section position="5" start_page="37" end_page="37" type="metho">
    <SectionTitle>
GLI TUTTO DALLA STANZA&amp;quot; (= remove everything from the room), is reported but, in
</SectionTitle>
    <Paragraph position="0"> this case, also some other correct sentences can be found by the parser (for example &amp;quot;COSTRUISCI UN TAVOLO QUADRATO&amp;quot; (build a square table).</Paragraph>
    <Paragraph position="1"> An efficient parser must also be able to solve other problems not strictly connected to a particular kind of input. In fact it should, of course, achieve fast operations; that requires the ability to minimize the number of alternative parses\[l\].</Paragraph>
    <Paragraph position="2"> Furthermore the parser should be designed in such a way as to satisfy the &amp;quot;generality&amp;quot; expectations; that is it should be easily adaptable to any semantic domain at least in the limited semantic domain cases. Since the parser results should be fol lowed by the execution of some operation in any practical application it is requi T red that it produced trusty results and, in particular, that it always included the right sentence interpretation within all the output ones.</Paragraph>
    <Paragraph position="3"> Finally, to allow a graceful dialogue with its users the parser must be able toana lyse also partial sentences (for example elliptical or fragmentary ones), thus ma Z king it possible to use naturally expressed sentences C2\].</Paragraph>
  </Section>
  <Section position="6" start_page="37" end_page="37" type="metho">
    <SectionTitle>
2. MAIN PARSER'S CHARACTERISTICS
</SectionTitle>
    <Paragraph position="0"> The main features of our parser, that permit to satisfy the above mentioned requirements, are the following: I) representation of the language in terms of a network whose elements are syntactic groups and syntactic features; 2) definition of a confidence measure of the recognition results and its extension also to groups of words (syntactic groups); 3) adoption of a recursive working strategy which anchors the parsing on the most reliable words in a first step and on the most reliable groups in a second one.</Paragraph>
    <Paragraph position="1"> We selected the furnishing of a living room as the discourse domain and we defined a vocabulary of a I16 words.</Paragraph>
    <Paragraph position="2"> This vocabulary, although limited, leads to a total number of over 10 5 possible sentences that include commands for constructing or moving pieces of furniture, assignment of labels, def~of unit lengths, inquiries about mutual distances~-e-TCT. null 2.1. Language representation We describe each sentence of the language by a sequence of syntactic groups. A sy~ tactic group is defined as a sentence part with a well precise semantic meaning. Often, but not necessary, a syntactic group corresponds to a classical grammatical object. For example, we defined the verb, the direct object, the location object, etc.</Paragraph>
    <Paragraph position="3"> In this way (see Fig. 2) each sentence of the langudge can be described by a sequence of some of these groups.</Paragraph>
    <Paragraph position="4"> On the whole we introduced only 9 groups; in our opinion this set of syntactic groups is enough to describe, at a syntactic level, all the possible sentences pe~ taining to this semantic environment.</Paragraph>
  </Section>
  <Section position="7" start_page="37" end_page="37" type="metho">
    <SectionTitle>
FLEXIBLE PARSING OF DISCRETELY UTTERED SENTENCES 39
</SectionTitle>
    <Paragraph position="0"/>
  </Section>
  <Section position="8" start_page="37" end_page="37" type="metho">
    <SectionTitle>
LANGUAGE REPRESENTATION
</SectionTitle>
    <Paragraph position="0"> Each syntactic group is, in turn, represented by a number of possible word sequences, or, mere precisely, of sequences of associated syntactic features. Figure 3 shows, for example, how the direct object is represented.</Paragraph>
  </Section>
  <Section position="9" start_page="37" end_page="37" type="metho">
    <SectionTitle>
DIRECT OBJECT REPRESENTATION
</SectionTitle>
    <Paragraph position="0"> It is important to notice ~n~L u,u Teature can represent mere than one word and every new word does not always need a new feature definition. So the present vocabulary can be easily increased to a certain degree within the semantic domain, with out any change in the grammar.</Paragraph>
    <Section position="1" start_page="37" end_page="37" type="sub_section">
      <SectionTitle>
2.2. Reliability evaluation
</SectionTitle>
      <Paragraph position="0"> The doubts connected with the vocal input suggested the need for a tool that measured the goodness of each word recognition. To this purpose a method to evaluate the reliability of recognition results was defined \[3\].</Paragraph>
    </Section>
  </Section>
  <Section position="10" start_page="37" end_page="37" type="metho">
    <SectionTitle>
8 I
</SectionTitle>
    <Paragraph position="0"> In this way, as described below, the most reliable words of the sentence can be se lected and the parser anchored to them. The same reliability score is also used to evaluate the syntactic groups found and to decide which, among alternative groups, is the most probable one.</Paragraph>
    <Section position="1" start_page="37" end_page="37" type="sub_section">
      <SectionTitle>
2.3 Island driven workin 9 strategy
</SectionTitle>
      <Paragraph position="0"> All the operations of the parser are centered around the concept of reliability score. In fact, in a first step, the parser anchors its analysis to the most reliable word of the sentence (that we named &amp;quot;guide word&amp;quot;) and searches, both to the right and to the left ofitforall the syntactic groups that include the features associated to the guide word. Each of these syntactic groups is named &amp;quot;island&amp;quot;.</Paragraph>
      <Paragraph position="1"> Not only the first word in the ordered list can be used for this aim, but sometimes also the second and the third ones are taken into consideration. For each island a cumulative reliability score, function of the single word scores, is computed.</Paragraph>
      <Paragraph position="2"> The same procedure is then applied to the remaining words untilthe whole sentence has been examined and there are no more guide words; at this point a lattice of island, possibly overlapping, is obtained.</Paragraph>
      <Paragraph position="3"> In a second step the parser, in an almost identical fashion as before, searches for the most reliable island (that we named &amp;quot;guide island&amp;quot;), anchoring to it the exploration of the language network to get a match with one of the possible sentences.</Paragraph>
      <Paragraph position="4"> When this is not possible, because of very unreliable recognition of a whole syntactic group, the partial sentence recovered is proposed in output together with a hypothesis about the missing constituent.</Paragraph>
      <Paragraph position="5"> At this point a module for graceful man-machine interaction could be activated, in order to obtain the needed information by meansofan appropriate dialogue.</Paragraph>
      <Paragraph position="6"> In addition there are some parameters, specifying the number of retained alternatives at various points of the parsing, that allow to control parser's performance both in terms of speed and confidence. These parameters allow the parser to work with different degrees of flexibility and so, they must be carefully selected, according to the application, i.e. according to the risk that can be tolerated when accepting an acoustically unclear sentence.</Paragraph>
    </Section>
  </Section>
  <Section position="11" start_page="37" end_page="37" type="metho">
    <SectionTitle>
FLEXIBL~ PARSING OF DISCRETELY UTTERED SENTENCES 4\]
3. RUNNING EXAMPLES
</SectionTitle>
    <Paragraph position="0"> In Fig. 5 the main steps of the analysis of a particular sentence are summarized.</Paragraph>
    <Paragraph position="1"> To make the comprehension easier we report a simulated english example that corresponds to a real italian sentence processed by the parser.</Paragraph>
  </Section>
  <Section position="12" start_page="37" end_page="37" type="metho">
    <SectionTitle>
INTO THE ROOM ,31 J ON THE TABLE ,11
INTO THE ROOM
</SectionTitle>
    <Paragraph position="0"> The input sentence is: PUT THE ROUND TABLE INTO THE ROOM. In the first step, starting respectively from the I st, 2 nd and 3rd guide word (PUT, ROUND, THE), the parser finds some possible islands with associated reliability score. In a second step, starting from the guide island the parser searches a match between a path in the language network and the islands. The final result is the correct interpretation of the sentence even if there were three recognition errors.</Paragraph>
    <Paragraph position="1"> Sometimes the parser outputs are not univocal as in the previous example. In fact, if the reliability score of a whole island is too low, the parser provides an output in which, instead of a detailed word-by-word interpretation, an hypothesis about the type of the missing syntactic group appears as shown in the example below: null PUT &lt; direct object &gt; IN THE MIDDLE OF THE ROOM If, on the contrary, tnere are two or more words with approximately the same relia bility score and the same syntactic role, then the parser supplies in the output those alternatives with their associated reliability scores and the whole decision will be deferred to a following pragmatic module or dialogue component. For example we can have an output like this one:</Paragraph>
  </Section>
  <Section position="13" start_page="37" end_page="37" type="metho">
    <SectionTitle>
PUT THE I TABLE CHAIR .32.29 I NUMBER TWO IN FRONT OF THE DOOR
4. RESULTS
</SectionTitle>
    <Paragraph position="0"> The present parser has been tested on a'set of 50 sentences spoken by three different speakers (two males and one female). A poor word recognizer was adopted in or der to stress parser's capabilities. We compare in table l the parser performanceT and those of the recognizer a~one. For each speaker the first column reports the percent of success of the recognizer alone, i.e. how many times all the words of a sentence were in the first position. The second column reports the percent of succes of the parser (i.e. how many times the parser was able to interpret the sentence). null Each row corresponds to a case in which there are, respectively, none,l, 2, and 3 lost islands whose reliability was not sufficient to take any decision.</Paragraph>
    <Paragraph position="1"> 42 L. BORGHESI and C. FAVARETO We want to notice that for the I st speaker the parser locates the correct sentence in the 92% of the cases and achieves a 96% correct interpretation if it assumes that there is one lost island. For the 2 nd speaker these values increase more slow ly because of a very unreliable input (10% of success for the recognizer).</Paragraph>
    <Paragraph position="2"> However the main result is that the parser never took a decision that did not contain the correct interpretation.</Paragraph>
    <Paragraph position="4"/>
  </Section>
class="xml-element"></Paper>