Levels Of Semantic Processing Research Articles

The application of morpho-syntactic language processing to effective phrase matching

The application of automatic natural language processing techniques to the indexing and the retrieval of text information has been a target of information retrieval researchers for some time. Incorporating semantic-level processing of language into retrieval has led to conceptual information retrieval, which is effective but usually restricted in its domain. Using syntactic-level analysis is domain-independent, but has not yet yielded significant improvements in retrieval quality. This paper describes a process whereby a morpho-syntactic analysis of phrases or user queries is used to generate a structured representation of text. A process of matching these structured representations is then described that generates a metric value or score indicating the degree of match between phrases. This scoring can then be used for ranking the phrases. In order to evaluate the effectiveness or quality of the matching and scoring of phrases, some experiments are described that indicate the method to be quite useful. Ultimately the phrasematching technique described here would be used as part of an overall document retrieval strategy, and some future work towards this direction is outlined.

Do children with learning disabilities outgrow deficits in selective auditory attention? Evidence from dichotic listening in adults with learning disabilities.

Previous research suggests that children with learning disabilities (LD) evidence similar degrees of cerebral lateralization as nondisabled children, but they perform at a degraded level due to deficits in selective attention. The developmental lag hypothesis predicts that these deficits should attenuate in children with LD with maturity. The present study examined this notion by evaluating dichotic listening ability using both free recall and directed attention conditions in 24 adults with LD and their age- and sex-matched controls. Dichotic syllables, words, and sentences were employed. The results indicated similar degrees of cerebral lateralization but significantly degraded levels of accuracy in the adults with LD in the free recall and directed attention conditions across all levels of semantic processing. Thus, it would appear that adults with LD evidence similar deficits in lateralized selective auditory linguistic processing as found in children with LD.

Levels of selection and capacity limits.

Subjects performed a visual target-detection task in eight experiments. We examined the effects of word relevancy (word in relevant or irrelevant location) and display load (1-4 words) on physical, semantic, and controlled processing of nontargets. Interwoven with the detection task was a test-word identification task that was used to measure priming potency of nontargets. Physical and semantic levels of processing were measured in terms of identity and semantic priming, respectively. Nontarget primes were repeated as test words in identity priming. Nontarget primes were semantic associates of test words in semantic priming. Controlled processing of nontargets was measured in terms of recognition memory on a subsequent test. All measures increased with word relevancy and decreased with display load. The priming effects remained intact even when word presentation was speeded up and controlled processing was sharply curtailed. The data indicate that all levels of processing are selective and capacity limited.

Need metaphoric comprehension take longer than literal comprehension?

To determine if metaphoric comprehension necessarily depends upon a more complex process than literal comprehension, 120 subjects in six different experiments were asked to code a series of sentences into one of the following logical sentence categories: analytic, synthetic, contradictory, anomalous, and metaphoric. Prior to this task, all subjects were given practice in learning to code examples of each of the various categories. Results of the present set of experiments revealed few systematic differences among the various categories in RT and a high degree of consistency in coding patterns across both the learning and RT phases of the experiments. Although procedural variations introduced for purposes of control produced clear and easily understandable changes in RT, coding patterns were essentially unchanged across the various experiments, suggesting that none of the categories could be considered as logically more basic than any other. The results were discussed in terms of contemporary semantic theory, where it was suggested that metaphoric recognition need not be conceptualized as depending on a more inferential level of semantic processing than literal recognition and that future theories must be more context-sensitive than those currently in fashion.

An Event-Related Potential Analysis of Coding Processes in Human Memory

This chapter discusses an event-related potential (ERP) analysis of coding processes in human memory. The endogenous ERPs have been shown to vary in the context of learning and memory tasks. For example, several lines of evidence indicate that the contingent negative variation (CNV) and P300 components reflect the processes of acquisition and retention in short term memory situations. One question to emerge from these results is the extent to which ERP changes can be related to retention over the longer term, i.e., in situations where memory is not tested immediately on every trial. It would be of interest to observe the extent to which ERPs reflect the level of perceptual processing at acquisition, and how such processing manifests itself in memory performance. Phonemic encodings yield intermediate performance. An ancillary finding is that words yielding a “yes” response to a query are better remembered than those yielding a “no” response in phonemic and semantic levels of processing. These effects have been obtained in a number of other experiments and represent a robust behavioral phenomenon with which to test the utility of ERPs as indices of learning and memory.

Melton and Irwin Updated: Retroactive Inhibition and Processing Level

Summary Two hundred sixty-five male and female undergraduates incidentally learned a mixed list of paired associates in an AB, AD and AB, CD transfer paradigm. Two levels of semantic processing were factorially varied across original and interpolated pairs. Recall of original pairs was shown to be directly related to the processing depth of original pairs in both paradigms and inversely related to the processing depth of interpolated pairs in the AB, AD paradigm but not in the AB, CD paradigm. Twenty additional Ss intentionally learned and recalled the same list and showed significant retroactive inhibition for original pairs in the AB, AD paradigm. The findings indicate that interference does not occur in a mechanical fashion, but is sensitive to Ss' level of perceptual analysis of the pairs.

Acoustic and semantic interference effects in words and pictures

Interference effects for pictures and words were investigated using a probe-recall task. Word stimuli showed acoustic interference effects for items at the end of the list and semantic interference effects for items at the beginning of the list, similar to results of Kintsch and Buschke (1969). Picture stimuli showed large semantic interference effects at all list positions with smaller acoustic interference effects. The results were related to latency data on picture-word processing and interpreted in terms of the differential order, probability, and/or speed of access to acoustic and semantic levels of processing. A levels of processing explanation of picture-word retention differences was related to dual coding theory. Both theoretical positions converge on an explanation of picture-word retention differences as a function of the relative capacity for semantic or associative processing.