Showing papers on "Phrase published in 1987"

Experiments in Automatic Phrase Indexing For Document Retrieval: A Comparison of Syntactic and Non-Syntactic Methods

Abstract: In order for an automatic information retrieval system to effectively retrieve documents related to a given subject area, the content of each document in the system''s database must be represented accurately. This study examines the hypothesis that better representations of document content can be constructed if the content analysis method takes into consideration the syntactic structure of document and query texts. Two methods of automatically generating phrases for use as content indicators have been implemented and tested experimentally. The non-syntactic (or statistical) method is based on simple text characteristics such as word frequency and the proximity of words in text. The syntactic method uses augmented phrase structure rules (production rules) to selectively extract phrases from parse trees generated by an automatic syntactic analyzer. Experimental results show that the effect of non-syntactic phrase indexing is inconsistent. For the five collections tested, increases in average precision ranged from 22.7% to 2.2% over simple, single term indexing. The syntactic phrase indexing method was tested on two collections. Precision figures averaged over all test queries indicate that non-syntactic phrase indexing performs significantly better than syntactic phrase indexing for one collection, but that the difference is insignificant for the other collection. More detailed analysis of individual queries, however, indicates that the performance of both methods is highly variable, and that there is evidence that syntax-based indexing has certain benefits not available with the non-syntactic approach. Possible improvements of both methods of phrase indexing are considered. It is concluded that the prospects for improving the syntax-based approach to document indexing are better than for the non-syntactic approach. The PLNLP system was used for syntactic analysis of document and query texts, and for implementing the syntax-based phrase construction rules. The SMART information retrieval system was used for retrieval experimentation.

Automatic phrase indexing for document retrieval

Abstract: An automatic phrase indexing method based on the term discrimination model is described, and the results of retrieval experiments on five document collections are presented. Problems related to this non-syntactic phrase construction method are discussed, and some possible solutions are proposed that make use of information about the syntactic structure of document and query texts.

The Use of Verbal and Imagery Mnemonics in Second-Language Vocabulary Learning

Abstract: This paper provides a critical look at the use of mnemonic associations in vocabulary learning. The paper begins with a definition of mnemonic devices—that is, techniques for converting material to be learned into a form that makes it easier to learn and remember—and focuses on verbal and imagery mnemonics, whereby a word, a phrase, or a sentence and visual imagery serve as mediator between what is known and what is to be learned. Particular attention is given to the keyword approach, in which there is both an acoustic link between a native-language word and the second-language word, and an image of the keyword interacting with the native-language word or phrase. Contentions are discussed concerning both the learning of words through verbal and imagery mnemonics and their subsequent retrieval. Attention is also given to research issues in need of investigation.

Government and tonal phrasing in Papago

Abstract: The intonation contour of a Papago sentence is entirely predictable on the basis of its surface syntactic structure. It consists of a sequence of instances of the pattern (L)HL. For example, one Papago sentence may manifest a single (L)HL pattern, as in (I):while another will show repetitions of the canonical tonal shape, as in (2):Our task in this paper will be to characterise the distribution of the (L)HL pattern in Papago sentences. Our analysis is that the phonological representation of a sentence of Papago consists of a sequence of one or more tonal phrases, and that (L)HL is the pattern assigned to a tonal phrase. This tonal phrasing - i.e. the beginnings and ends of individual tonal phrases - will be indicated by parentheses in the tonal tier, as in (1) and (2). The association of the (L)HL pattern within each tonal phrase is captured by simple rules of the phonology, to be described in §2. Thus the description of the syntax of Papago intonation contours will centre on the tonal phrasing itself, and on the manner in which it is determined with respect to surface syntactic structure.

Fixed text speaker verification method and apparatus

Abstract: Speaker verification is performed by computing principal components of a fixed text statement comprising a speaker identification code and a two-word phrase, and principal spectral components of a random word phrase. A multi-phrase strategy is utilized in access control to allow successive verification attempts in a single session, if the speaker fails initial attempts. Based upon a verification attempt, the system produces a verification score which is compared with a threshold value. On successive attempts, the criterion for acceptance is changed, and one of a number of criteria must be satisfied for acceptance in subsequent attempts. A speaker normalization function can also be invoked to modify the verification score of persons enrolled with the system who inherently produce scores which result in denial of access. Accuracy of the verification system is enhanced by updating the reference template which then more accurately symbolizes the person's speech signature.

Osteochondritis dissecans 1887-1987. A centennial look at Konig's memorable phrase

Abstract: In 1887, Konig claimed that loose bodies in the knee joints of young persons had three causes: (1) very severe trauma; (2) lesser trauma causing contusion and necrosis; and (3) minimal trauma acting on an underlying lesion-for which he suggested the name osteochondritis dissecans. His thesis has stood the test of time. We still confuse the second and third categories but osteochondritis dissecans has been identified as an ossification defect.

The Theory of Parameters and Syntactic Development

Abstract: Within recent theories of Generative Grammar, UG has taken the form of a parameterized system. The parameters of UG express the limited range of variation which is permitted with respect to a set of core principles. For example, the X-Bar schemata given in (1) specifies that the categorial rules must conform to an ‘endocentric requirement’; that is, each phrase must contain a lexical head of its own feature specification. However, languages may vary as to the linear position of the head with respect to its complements (represented by ...) (Stowell, 1981).

Incremental interpretation and combinatory categorial grammar

Abstract: The paper is organised around a particular problem of ana­ phoric definite reference. A solution to this problem requires sen­ tence processing which allows semantic representations to be evaluated incrementally, as a phrase is read from left to right. The paper shows how a proposal made by Mellish (1985), to regard incremental semantic evaluation as a constraint satisfaction task, can provide a natural solution to the problem of reference. The incorporation of Combinatory Categorial Grammar allows a straightforward relationship to be stated between the incremen­ tally assembled syntactic and semantic representations.

Theoretical notes on “Parallel models of associative memory”

Abstract: A phrase which has become fashionable in discussing neuropsychological functioning is “parallel distributed processing”. The meaning and implica- tions of this phrase are examined in the papers edited by Hinton and Ander- son (1981) in “Parallel models of associative memory” (henceforth, PMAM). Since the book was published the theoretical ideas discussed in it have steadily increased in popularity and influence. Indeed, in retrospect, it appears as though the book set out the beginnings of an alternative framework for understanding neuropsychological functioning. Given this, and the current interest in cognitive neuropsychology, it is appropriate to examine the book and the new framework critically.

Where do Phrases Come from: Some Preliminary Experiments in Connectionist Phrase Generation

Abstract: Some of the assumptions that were inherent in many of our early models of language generation (assumptions such as the unidirectional flow of information from conceptualization to verbalization, the serial processing of semantic, lexical, syntactic and phonetic information, and the importance of the word as the lexical unit) are coming under closer scrutiny within the symbolic modeling paradigm. At the same time, interest in another paradigm of cognitive modeling, one that has been variously referred to as neural network modeling, parallel distributed processing, and connectionism, has been growing. Certain characteristics of connectionist models, in particular their broad bandwidth information channels which permit parallel, multidirectional information flow, seem especially relevant to current language generation problems. Other useful properties of these models, such as their content addressable memory capacity, their ability to respond to novelty, their capacity to degrade gracefully and the existence of automatic learning procedures, would contribute to the solutions of other language generation problems. This paper gives an overview of the connectionist paradigm and a brief summary of some of the language processing research being done under it. It then describes the results of some experiments aimed at training connectionist networks to perform two language generation tasks. In one study a network was trained to produce appropriate words given a set of semantic attributes, and in another study a network was trained to produce syntactically well-formed phrases given the unordered words. The networks finally did achieve their intended behaviors, and in doing so they exhibited evidence of having acquired linguistic knowledge of synonymy relations and collocation relations in addition to proper semantic associations. Many of the errors made by the networks suggested revisions to network configurations and encoding schemes, and the effects of those revisions are also briefly described.

Wa and the WH phrase

Abstract: A pin tumbler cylinder lock having a plug eccentrically positioned within a cylindrical bore of a housing. Surrounding the plug are longitudinally aligned sleeve segments that together therewith form a composite plug unit that can be rotated in the housing bore by a proper key. Formed on the outer surface of each sleeve segment coincident with pinways therein are recesses that are spaced from the recesses in adjacent sleeve segments. Additional internal recesses are formed in the inner surface of at least one of the sleeve segments closely adjacent a pinway extending therethrough. The internal recesses receive the tumbler pin when it is forced upwardly simultaneously with the application of torque on the plug. In one embodiment, the internal recess is a single recess coincident with the pinway and extending circumferentially therefrom. In another embodiment, the internal recess comprises a pair of recesses located on opposite sides of the pinway and spaced circumferentially therefrom.

Method and apparatus of rejecting false hypotheses in automatic speech recognizer systems

Abstract: An automatic speech recognition system employs parallel syntaxes with a first syntax operative to compare keyword templates with incoming speech over a given time interval and with a second syntax in parallel with the first and operative to compare filler templates with incoming speech over the same interval. Based on the best comparisons in each syntax a likelihood probability ratio is computed and compared against a selected threshold for determining whether the speech contains a valid phrase or keyword as compared to an undesirable utterance.

On defining the intonational phrase: evidence from Slave

Abstract: Syntactic juncture has been a topic of interest in phonological theory in recent years. One major issue addressed in the study of syntactic juncture is how to predict from syntactic structure the domains of phrase-level rules of the phonology, or prosodic structure. Many, including Selkirk (1978, 1984, 1986), Nespor & Vogel (1982) and Hayes (1984), propose that utterances are organised in a prosodic hierarchy, determined by but not isomorphic to syntactic structure. In work by these authors, algorithms for determining the relationship between syntactic structure and prosodic structure have been proposed, leading to a deeper understanding of prosodic phrasing.

An Acoustic Study on the Differences of Contrastive Emphasis Between Native and Non-Native Spanish Speakers.

Abstract: Since intonation is a suprasegmental element of phonology, it does not receive as much attention in a language class as many other areas of language teaching. This is unfortunate because correct intonation is a vital part of being understood. John Dalbor (1969:223) has observed, "Intonation is so important in language that it not only controls grammatical meaning to a large extent, but it also conveys the attitude of the speaker even more effectively than the words he chooses." Bowen (1956:33) states that improper intonation in moments of high emotion may cause the non-native speaker of Spanish to sound angry or disgusted. The purpose of this study is to focus on different ways to express contrastive emphasis in Spanish and English to determine how the range of pitch of advanced non-native speakers of Spanish compares with that of native Spanish speakers. This will be done by acoustically comparing their Spanish intonation patterns with those of native Spanish speakers and also by comparing them with English intonation patterns of native English speakers. There are some often used terms that may need to be more closely defined for this study. These terms are: contrast, emphasis, intensity, intonation, pitch, and stress. Webster's dictionary defines contrast as showing differences when comparing. Emphasis is defined as a special stress given to a syllable, word, etc. in speaking. Therefore, one of the ways to contrast something is to emphasize it with a special stress. Ladefoged defines stress as "the extra use of respiratory energy during a syllable" (1975:285). He also defines intensity as the amount of acoustic energy in a sound (1975:282). In other words, stress is extra perceived energy and intensity is extra energy that can be acoustically measured. Intonation is the pattern of pitch changes that occur during a phrase (Ladefoged 1975:252, Quilis and Fernaindez 1973:164). Pitch is the "auditory property of a sound that enables a listener to place it on a scale going from low to high, without considering the acoustic properties such as the frequency of the sound" (Ladefoged 1975:284). This study will key on contrasts made through the use of pitch and intensity. The term "normal tone frequency" will be defined as pitch of a subject's voice at the beginning of a phrase. "Contrasted tone frequency" is defined as the pitch of the voice in contrasted syllables. Likewise, "normal intensity" will be the intensity at the beginning of a phrase while "contrasted intensity" will be the intensity of contrasted syllables.

Towards a discourse-based model of English sentence intonation

Abstract: English non-expressive, declarative sentence intonation is examined in a discourse context. A rule system, geared to a text-to-speech context is developed which assigns sentence prominences related to information focus. The realization of these prominences with respect to the phonetic parameter of pitch (F0) is also described with a pitch generating algorithm. The dynamics of information focus in English is shown to be accountable for by a hierarchy of grammatical functions interacting with a contextual notion of coreference. Projection of focal prominence is effected by a subroutine which assigns prominence either to the head or to the modifier of the constituent realizing a particular grammatical function, depending on the coreferential status of the associated lexical items. The model assumes three degrees of prominence associated with the three grammatical functions (Subject, Predicate, Predicate complement). These are realized by varying heights of F0 peaks described in relation to the baseline of a phonological 'grid' characterizing the speaker's F0 range and declination constants. A fourth degree of prominence is also shown to be required on prefocal lexical heads. The notion of contrastive prominence conditioned by syntactic parallelism and identity relations is examined and preliminary rules are formulated for assigning and synthesizing the associated F0 contours. The problem of noncontraction of the copula is also related to the issue of contrastive prominence. The status of the notion of 'phonological phrase' is examined in the light of a grammar which includes a focus assigning component and a number of phonological processes (e.g. 'rhythm rule') are reanalyzed. (Less)

Kana-kanji converting device

Abstract: PURPOSE:To convert a KANA (Japanese syllabary) character-string into a KANA-KANJI (Chinese character) mixed sentence by providing a HIRAGANA (cursive Japanese syllabary) indefinite phrase and KATAKANA (square form of Japanese syllabary) indefinite phrase generating means for an indefinite word written with a space between words, a simple clause, and an indefinite phrase word learning means to display, select and define those and register automatically an independent word in a dictionary. CONSTITUTION:Under the control of a CPU, the inputted KANA character-string is converted into the KANJI by a converting device 8, and a candidate is displayed and selected by a clause candidate displaying part 3 and a clause candidate selecting part 4, and the length of a clause is changed 6. The KANA is designated 7, and the candidate of the clause is fixed 5. In the converting device 8, a KANA character is received by a KANA character acquiring part 9, and a clause writing part 10 with a space between words refers to a word dictionary 11, and inter-word connection is approved 14. The indefinite phrase generating part 13 stores the indefinite phrase from the indefinite word character-string and the clause candidate in a clause candidate storage part 16, and outputs them by a clause unit from a clause candidate outputting part 15. The change of sentence length is reanalyzed by a clause reanalyzing part 12, and the indefinite phrase is generated, and the rest is written with a space between words into the clause. When the KANA is designated, the analyzing part 12 determines KANA priority, and the defined clause candidate is dissolved into the independent word and an adjunctive word by a receiving part 17, and the dictionary is made learn them.

Orthography as a pronunciation resource

Abstract: Of the newest trends in English pronunciation teaching-the emphases on fast-speech phenomena, phrase rhythm, word stress, and the use of orthography as a learning tool-only this last is calculated to produce independent, self-sufficient learners, capable of increasing in oral proficiency outside the classroom, away from textbooks, and beyond the influence of teachers. This article draws together and illustrates the most compelling reasons for teachers to introduce spelling information into their pronunciation work. Highlighted throughout are two points: (1) the significant benefits to students from a systematic exposure to spelling-based prediction patterns, and (2) the nature of learner-orientated rules of pronunciation, often less familiar to teachers than rules of grammar.

Strategies in language acquisition: learning phrases in context

Abstract: How is language acquired by people, and how can we make computers simulate language acquisition? Although current linguistic models have investigated extensively parsing and generation, so far, there has been no model of learning new lexical phrases from examples in context. We have identified four issues in language acquisition. (a) How can a phrase be extracted from a single example? (b) How can phrases be refined as further examples are provided? (c) How can the context be incorporated as part of a new phrase? (d) How can acquired phrases be used in parsing and in generation? In solving this problems, we have established three theoretical points. (a) We have shown how a dynamic lexicon is structured as a phrasal hierarchy. (b) We have constructed strategies for learning phrases. (c) We have constructed a parsing mechanism which can operate even in the presence of lexical gaps. The program RINA has incorporated these elements in modeling a second-language speaker who augments her lexical knowledge by being exposed to examples in context.

Document retrieving and displaying method

Abstract: PURPOSE: To automatically display a corresponding sentence or a part of it in juxtaposition only designating character string by combining a step to specify a designated phrase, a step to retrieve an object designated in the designated phrase, and a step in which a document or a part of it that includes the object of designation is displayed in an appropriate position on the screen CONSTITUTION: When an operator designates a designated phrase 52 by executing a position-designation by using a transparent tablet 6a provided on top of the screen of a cathode-ray tube 5 and a mouse/cursor 6c that moves when a mouse 6b is moved, the system extracts the designated phrase and specifies it out of the designated position and the document data that originates the display Then a part corresponding to thus specified designated phrase is searched in the system, and displayed on the screen in a proper disposition As a result, a part of document including the designated object 54 for the designated phrase 52 is displayed in juxtaposition in a frame 53 COPYRIGHT: (C)1989,JPO&Japio

Introduction: syntactic influences on phonological rules

Abstract: Phonological theory asks how phonological shapes are assigned to linguistic units. In this issue of the Phonology Yearbook we bring together examinations of the role that syntactic structure plays in this assignment. The focus is therefore on phenomena that have been labelled, in one framework or another, as external sandhi (as opposed to internal sandhi), phrase phonology (as opposed to word phonology), or postlexical phonology (as opposed to lexical phonology). We have organised the eleven articles into three groups. In the first group are four articles addressing the issue of what domains phonological rules apply in. There is no question that there are phonological rules applying in purely phonological, or prosodic, domains (P2 rules, in the terminology of Kaisse i985); rules traditionally classified as allophonic, in particular, are all of this sort. There is also no question that prosodic domains can be related to syntactic (and morphological) domains in systematic, though often rather complex ways, so that syntactic structure on many occasions influences the applicability of P2 rules but indirectly, via the mediation of the principles relating syntactic domains to prosodic domains. What is in question in our first group of articles is whether there are phonological rules subject to conditions that are best stated directly in terms of syntactic domains (Pi rules, in Kaisse's terminology). Odden, for instance, defends a direct-reference approach to the analysis of some phenomena occurring within maximal projections in Kimatuumbi, arguing that a mediated-reference approach couched in terms of prosodic domains is inadequate. The three articles in our middle group explore a new approach to mediated-influence analyses of phonological rules that are sensitive to syntactic information; this is the end-based theory introduced by Selkirk

How to terminate a phrase. An analysis-by-synthesis experiment on the perceptual aspect of music performance

Abstract: How to terminate a phrase. An analysis-by-synthesis experiment on the perceptual aspect of music performance

Voice recognition system

Abstract: PURPOSE: To perform accurate recognition by comparing the recognition result of respoken input with the recognition result before the respoken input, performing dictionary retrieval based upon the same parts of both the recognition, and calculating the likelihood between a selected word or phrase and the input word. CONSTITUTION: When a recognition result is corrected, the recognition result of the respoken input and the recognition result before the respoken input are compared with each other to detect the same parts of both the recognition results, dictionary retrieval is performed with only the data of the same parts, and the likelihood between the selected word or phrase and respoken input word is calculated. To avoid the word rerecognition of a phrase and the recognition of a word as a paragraph at the time of the respoken input, a voice recognition part 1 performs word recognition and paragraph recognition in parallel, the recognition results of the word and paragraph are found, and the results of paragraph recognition processing are displayed on a display device 6 in the decreasing order of likelihood according to both the recognition results. Consequently, accurate and speedy recognition can be performed.

Speech recognition apparatus and methods

Abstract: Speech recognition apparatus has a vocabulary store 2 containing pattern templates or other acoustic information, grammar and collocation information about each word that can be recognised. An acoustic recogniser 30 compares the output of a microphone 1 with the pattern templates in the store 2 to produce signals representing the most likely words spoken and their grammar and collocation. The grammar information is used to check that the words identified by the recogniser are compatible with each other. Grammar information can also be used to restrict the words in the store with which following words are compared by the recogniser 30. A semantics unit 40 compares collocation information of identified words with other identified words in a phrase to check that they are compatible. If identified words are incompatible in grammar or collocation, alternative, less likely words are selected for comparison. When a word is identified with sufficient probability that is compatible in grammar and collocation, an output is generated to a utilisation device 5. If no word is identified with sufficient certainty, a feedback signal is produced to prompt repetition of the word.