Lexical Cues Research Articles

Comparing Lexical Cues in Listener Processing of Dysarthria and Speech in Noise.

Purpose The frequency of a word and its number of phonologically similar neighbors can dramatically affect how likely it is to be accurately identified in adverse listening conditions. This study compares how these two cues affect listeners' processing of speech in noise and dysarthric speech. Method Seven speakers with moderate hypokinetic dysarthria and eight healthy control speakers were recorded producing the same set of phrases. Statements from control speakers were mixed with noise at a level selected to match the intelligibility range of the speakers with dysarthria. A binomial mixed-effects model quantified the effects of word frequency and phonological density on word identification. Results The model revealed significant effects of word frequency (b = 0.37, SE = 0.12, p = .002) and phonological neighborhood density (b = 0.40, SE = 0.12, p = .001). There was no effect of speaking condition (i.e., dysarthric speech vs. speech in noise). However, a significant interaction was observed between speaking condition and word frequency (b = 0.26, SE = 0.04, p < .001). Conclusions The model's interactions indicated that listeners were more strongly influenced by the effects of word frequency when decoding moderate hypokinetic dysarthria as compared to speech in noise. Differences in listener reliance on lexical cues may have important implications for the selection of communication-based treatment strategies for speakers with dysarthria.

Rational Adaptation in Using Conceptual Versus Lexical Information in Adults With Aphasia.

The information theoretic principle of rational adaptation predicts that individuals with aphasia adapt to their language impairments by relying more heavily on comparatively unimpaired non-linguistic knowledge to communicate. This prediction was examined by assessing the extent to which adults with chronic aphasia due to left-hemisphere stroke rely more on conceptual rather than lexical information during verb retrieval, as compared to age-matched neurotypical controls. A primed verb naming task examined the degree of facilitation each participant group received from either conceptual event-related or lexical collocate cues, compared to unrelated baseline cues. The results provide evidence that adults with aphasia received amplified facilitation from conceptual cues compared to controls, whereas healthy controls received greater facilitation from lexical cues. This indicates that adaptation to alternative and relatively unimpaired information may facilitate successful word retrieval in aphasia. Implications for models of rational adaptation and clinical neurorehabilitation are discussed.

The visual and semantic features that predict object memory: Concept property norms for 1,000 object images

Humans have a remarkable fidelity for visual long-term memory, and yet the composition of these memories is a longstanding debate in cognitive psychology. While much of the work on long-term memory has focused on processes associated with successful encoding and retrieval, more recent work on visual object recognition has developed a focus on the memorability of specific visual stimuli. Such work is engendering a view of object representation as a hierarchical movement from low-level visual representations to higher level categorical organization of conceptual representations. However, studies on object recognition often fail to account for how these high- and low-level features interact to promote distinct forms of memory. Here, we use both visual and semantic factors to investigate their relative contributions to two different forms of memory of everyday objects. We first collected normative visual and semantic feature information on 1,000 object images. We then conducted a memory study where we presented these same images during encoding (picture target) on Day 1, and then either a Lexical (lexical cue) or Visual (picture cue) memory test on Day 2. Our findings indicate that: (1) higher level visual factors (via DNNs) and semantic factors (via feature-based statistics) make independent contributions to object memory, (2) semantic information contributes to both true and false memory performance, and (3) factors that predict object memory depend on the type of memory being tested. These findings help to provide a more complete picture of what factors influence object memorability. These data are available online upon publication as a public resource.

Sarcasm Detection in Tweets: A Feature-based Approach using Supervised Machine Learning Models

Sarcasm (i.e., the use of irony to mock or convey contempt) detection in tweets and other social media platforms is one of the problems facing the regulation and moderation of social media content. Sarcasm is difficult to detect, even for humans, due to the deliberate ambiguity in using words. Existing approaches to automatic sarcasm detection primarily rely on lexical and linguistic cues. However, these approaches have produced little or no significant improvement in terms of the accuracy of sentiment. We propose implementing a robust and efficient system to detect sarcasm to improve accuracy for sentiment analysis. In this study, four sets of features include various types of sarcasm commonly used in social media. These feature sets are used to classify tweets into sarcastic and non-sarcastic. This study reveals a sarcastic feature set with an effective supervised machine learning model, leading to better accuracy. Results show that Decision Tree (91.84%) and Random Forest (91.90%) outperform in terms of accuracy compared to other supervised machine learning algorithms for the right features selection. The paper has highlighted the suitable supervised machine learning models along with its appropriate feature set for detecting sarcasm in tweets.

Learning Lexical Subspaces in a Distributional Vector Space

In this paper, we propose LexSub, a novel approach towards unifying lexical and distributional semantics. We inject knowledge about lexical-semantic relations into distributional word embeddings by defining subspaces of the distributional vector space in which a lexical relation should hold. Our framework can handle symmetric attract and repel relations (e.g., synonymy and antonymy, respectively), as well as asymmetric relations (e.g., hypernymy and meronomy). In a suite of intrinsic benchmarks, we show that our model outperforms previous approaches on relatedness tasks and on hypernymy classification and detection, while being competitive on word similarity tasks. It also outperforms previous systems on extrinsic classification tasks that benefit from exploiting lexical relational cues. We perform a series of analyses to understand the behaviors of our model. 1 Code available at https://github.com/aishikchakraborty/LexSub .

Acoustic-Prosodic and Lexical Cues to Deception and Trust: Deciphering How People Detect Lies

Humans rarely perform better than chance at lie detection. To better understand human perception of deception, we created a game framework, LieCatcher, to collect ratings of perceived deception using a large corpus of deceptive and truthful interviews. We analyzed the acoustic-prosodic and linguistic characteristics of language trusted and mistrusted by raters and compared these to characteristics of actual truthful and deceptive language to understand how perception aligns with reality. With this data we built classifiers to automatically distinguish trusted from mistrusted speech, achieving an F1 of 66.1%. We next evaluated whether the strategies raters said they used to discriminate between truthful and deceptive responses were in fact useful. Our results show that, although several prosodic and lexical features were consistently perceived as trustworthy, they were not reliable cues. Also, the strategies that judges reported using in deception detection were not helpful for the task. Our work sheds light on the nature of trusted language and provides insight into the challenging problem of human deception detection.

Neural Correlates of Phonetic Adaptation as Induced by Lexical and Audiovisual Context.

When speech perception is difficult, one way listeners adjust is by reconfiguring phoneme category boundaries, drawing on contextual information. Both lexical knowledge and lipreading cues are used in this way, but it remains unknown whether these two differing forms of perceptual learning are similar at a neural level. This study compared phoneme boundary adjustments driven by lexical or audiovisual cues, using ultra-high-field 7-T fMRI. During imaging, participants heard exposure stimuli and test stimuli. Exposure stimuli for lexical retuning were audio recordings of words, and those for audiovisual recalibration were audio-video recordings of lip movements during utterances of pseudowords. Test stimuli were ambiguous phonetic strings presented without context, and listeners reported what phoneme they heard. Reports reflected phoneme biases in preceding exposure blocks (e.g., more reported /p/ after /p/-biased exposure). Analysis of corresponding brain responses indicated that both forms of cue use were associated with a network of activity across the temporal cortex, plus parietal, insula, and motor areas. Audiovisual recalibration also elicited significant occipital cortex activity despite the lack of visual stimuli. Activity levels in several ROIs also covaried with strength of audiovisual recalibration, with greater activity accompanying larger recalibration shifts. Similar activation patterns appeared for lexical retuning, but here, no significant ROIs were identified. Audiovisual and lexical forms of perceptual learning thus induce largely similar brain response patterns. However, audiovisual recalibration involves additional visual cortex contributions, suggesting that previously acquired visual information (on lip movements) is retrieved and deployed to disambiguate auditory perception.

Contributions of acoustic and lexical cue weighting to individual differences in speech perception

Past research suggests that individuals with weaker receptive language show increased reliance on lexical information for speech perception relative to individuals with stronger receptive language, which may reflect a difference in how acoustic and lexical cues are weighted during online speech processing. Here we investigated whether this relationship is the consequence of removing natural correlations between acoustic and lexical cues in speech input, as in the specific stimulus distributions used in previous research. Two groups of adult participants completed measures of receptive and expressive language ability in addition to a measure of lexical recruitment (i.e., a phonetic identification task to measure the Ganong effect). In the high conflict group, the stimulus input distributions put acoustic and lexical cues in high competition; in the low conflict group, this competition was reduced. The results showed that (1) the Ganong effect was larger in the low compared to the high conflict condition, (2) the Ganong effect was larger for those with weaker compared to stronger receptive language, and (3) the relationship between the Ganong effect and receptive language was not mediated by conflict. These results demonstrate increased reliance on lexical information among those with weaker receptive language, suggesting that listeners with weaker language abilities down-weight acoustic cues and rely more heavily on lexical knowledge even when natural acoustic and lexical cue correlations are maintained in the input.

Two-Level LSTM for Sentiment Analysis With Lexicon Embedding and Polar Flipping.

Sentiment analysis is a key component in various text mining applications. Numerous sentiment classification techniques, including conventional and deep-learning-based methods, have been proposed in the literature. In most existing methods, a high-quality training set is assumed to be given. Nevertheless, constructing a high-quality training set that consists of highly accurate labels is challenging in real applications. This difficulty stems from the fact that text samples usually contain complex sentiment representations, and their annotation is subjective. We address this challenge in this study by leveraging a new labeling strategy and utilizing a two-level long short-term memory network to construct a sentiment classifier. Lexical cues are useful for sentiment analysis, and they have been utilized in conventional studies. For example, polar and negation words play important roles in sentiment analysis. A new encoding strategy, that is, ρ -hot encoding, is proposed to alleviate the drawbacks of one-hot encoding and, thus, effectively incorporate useful lexical cues. Moreover, the sentimental polarity of a word may change in different sentences due to label noise or context. A flipping model is proposed to model the polar flipping of words in a sentence. We compile three Chinese datasets on the basis of our label strategy and proposed methodology. Experiments demonstrate that the proposed method outperforms state-of-the-art algorithms on both benchmark English data and our compiled Chinese data.

Lexical and contextual cue effects in discourse expectations: Experimenting with German 'zwar...aber' and English 'true/sure...but'

Existing literature shows that readers and listeners rapidly adjust their expectations about likely discourse continuations through discourse markers, as well as through other linguistic and extra-linguistic cues. However, it is unclear whether (i) the facilitative effects of various (extra-)linguistic cues differ in quality and (ii) whether the effects interact with one another in any principled manner. We conducted two self-paced reading experiments on concessive constructions in German and English wherein optional lexical and/or contextual cues appeared ahead of the concessive discourse marker. The results demonstrate that readers can use both types of cues to anticipate the upcoming discourse relation. Our study thus provides novel evidence for expectation-driven accounts of discourse processing and elucidates the functions of discourse signals. Furthermore, the results also show that the role that a type of cues plays is subject to cross-linguistic variation.

Syntactic Cues Help Disambiguate Objects Referred to With Count Nouns: Illustration With Malay Children.

Children employ multiple cues to identify the referent of a novel word. Novel words are often embedded in sentences and children have been shown to use syntactic cues to differentiate between types of words (adjective vs. nouns) and between types of nouns (count vs. mass nouns). In this study, we show that children learning Malay (N=67), a numeral classifier language, can use syntactic cues to perform even finer-grained disambiguation-between count nouns. The manipulation of congruence between lexical and syntactic cues reveals a clear developmental trajectory: while 5-year-olds use predominantly lexical cues, older children increasingly rely on syntactic cues, such that by 7years of age, they disambiguate between objects referred to with count nouns using syntactic rather than lexical cues.

Real-Time Interruption Management System for Efficient Distributed Collaboration in Multi-tasking Environments

Interruption dissemination in proactive systems remains a challenge for efficient human-machine collaboration, especially in real-time distributed collaborative environments. In this paper a real-time interruption management system (IMS) is proposed that leverages speech information, the most commonly used and available means of communication within collaborative distributed environments. The key aspect of this paper includes a proposed real-time IMS system that leverages lexical affirmation cues to infer the end of a task or task boundary as a candidate interruption time. The performance results show the proposed real-time lexical Affirmation Cues based Interruption Management System (ACE-IMS) outperforms the current baseline real-time IMS system within the existing literature. ACE-IMS has the potential of reducing disruptive interruptions without incurring excessive missed opportunities to disseminate interruptions by utilizing only the most frequently used mode of human communication: voice. Thereby, providing a promising new baseline to further the system development of real-time interruption management systems within the ever-growing distributed collaborative domain.

A Longitudinal Study of the Acquisition of the Polysemous Verb 打 dǎ in Mandarin Chinese

Lexical ambiguity abounds in languages and multiple one-to-many form-function mappings create challenges for language learners. This study extends the theoretical approaches to the acquisition of polysemy to the Mandarin verb 打 dǎ, which is highly polysemous and among the earliest verbs in child speech. It analyzes longitudinal naturalistic data of nine children (1;05–3;10) from two Mandarin child speech corpora to explore the developmental trajectory of different senses of 打 dǎ and the role of input. The results support a continuous derivational and restricted monosemy approach: children initially extract a core feature of 打 dǎ, but only apply it in a restricted way, reflected in a small number of senses in a limited set of semantic domains and syntactic frames, revealing an early preference for initial unambiguous form-meaning mappings. Mandarin-speaking children’s production mirrors the semantic and syntactic distribution of the input, supporting the usage-based approach to the acquisition of polysemy that meaning is derived from the confluence of lexical and syntactic cues in the usage patterns in the input. Our research is the first longitudinal study of the emergence and development of polysemous verbs in Mandarin and has pedagogical implications for teaching Mandarin as a second language.

Audiovisual and lexical cues do not additively enhance perceptual adaptation

When listeners experience difficulty in understanding a speaker, lexical and audiovisual (or lipreading) information can be a helpful source of guidance. These two types of information embedded in speech can also guide perceptual adjustment, also known as recalibration or perceptual retuning. With retuning or recalibration, listeners can use these contextual cues to temporarily or permanently reconfigure internal representations of phoneme categories to adjust to and understand novel interlocutors more easily. These two types of perceptual learning, previously investigated in large part separately, are highly similar in allowing listeners to use speech-external information to make phoneme boundary adjustments. This study explored whether the two sources may work in conjunction to induce adaptation, thus emulating real life, in which listeners are indeed likely to encounter both types of cue together. Listeners who received combined audiovisual and lexical cues showed perceptual learning effects similar to listeners who only received audiovisual cues, while listeners who received only lexical cues showed weaker effects compared with the two other groups. The combination of cues did not lead to additive retuning or recalibration effects, suggesting that lexical and audiovisual cues operate differently with regard to how listeners use them for reshaping perceptual categories. Reaction times did not significantly differ across the three conditions, so none of the forms of adjustment were either aided or hindered by processing time differences. Mechanisms underlying these forms of perceptual learning may diverge in numerous ways despite similarities in experimental applications.

Pragmatic Aspects of Discourse Production for the Automatic Identification of Alzheimer's Disease

Clinical literature provides convincing evidence that language deficits in Alzheimer's disease (AD) allow for distinguishing patients with dementia from healthy subjects. Currently, computational approaches have widely investigated lexicosemantic aspects of discourse production, while pragmatic aspects like cohesion and coherence, are still mostly unexplored. In this article, we aim at providing a more comprehensive characterization of language abilities for the automatic identification of AD in narrative description tasks by also incorporating pragmatic aspects of speech production. To this end, we investigate the relevance of a recently proposed set of pragmatic features extracted from an automatically generated topic hierarchy graph in combination with a complementary set of state-of-the-art features encoding lexical, syntactic and semantic cues. Experimental results on the DementiaBank corpus show an accuracy improvement from 82.6% to 85.5% in identifying AD patients when pragmatic features are incorporated to the set of lexicosemantic features. Nevertheless, these results are obtained relying on manual transcriptions, which strongly limits the applicability of computational analysis to clinical settings. Thus, in this work we additionally carry out an analysis of the errors introduced by a speech recognition system and the way in which they impact the performance of the proposed method. In spite of the high word error rates obtained on these data (∼40%), automatic AD identification accuracy decreased only to 79.7%, which is considered a remarkable result when compared with solutions based on manual transcriptions.

Interleaved lexical and audiovisual information can retune phoneme boundaries.

To adapt to situations in which speech perception is difficult, listeners can adjust boundaries between phoneme categories using perceptual learning. Such adjustments can draw on lexical information in surrounding speech, or on visual cues via speech-reading. In the present study, listeners proved they were able to flexibly adjust the boundary between two plosive/stop consonants, /p/-/t/, using both lexical and speech-reading information and given the same experimental design for both cue types. Videos of a speaker pronouncing pseudo-words and audio recordings of Dutch words were presented in alternating blocks of either stimulus type. Listeners were able to switch between cues to adjust phoneme boundaries, and resulting effects were comparable to results from listeners receiving only a single source of information. Overall, audiovisual cues (i.e., the videos) produced the stronger effects, commensurate with their applicability for adapting to noisy environments. Lexical cues were able to induce effects with fewer exposure stimuli and a changing phoneme bias, in a design unlike most prior studies of lexical retuning. While lexical retuning effects were relatively weaker compared to audiovisual recalibration, this discrepancy could reflect how lexical retuning may be more suitable for adapting to speakers than to environments. Nonetheless, the presence of the lexical retuning effects suggests that it may be invoked at a faster rate than previously seen. In general, this technique has further illuminated the robustness of adaptability in speech perception, and offers the potential to enable further comparisons across differing forms of perceptual learning.

Three-year-olds infer polite stance from intonation and facial cues

AbstractDespite the evidence that infants are sensitive to facial cues and prosody for the detection of emotion, we have contradictory evidence regarding the use of these cues by older preschool and school children when inferring both emotional and politeness stance. This study assessed preschool aged children’s sensitivity to intonational and facial cues signalling a speaker’s polite stance in requestive speech acts with controlled lexical and contextual materials. Thirty-six 3-year-old American English-speaking children performed a forced-choice decision task which investigated whether children at this age use pitch and/or facial cues to infer a speaker’s affective stance in either audio-only, visual-only or audio-visual presentation modalities, when lexical cues are controlled for. Results showed that (a) children at three years can infer a speaker’s polite stance equally well in all three conditions (audio-only, visual-only and audio-visual) and thereby (b) unlike previous research, in the present task both intonation and facial cues are equally strong cues in children’s understanding of a speaker’s polite stance in requestive speech acts. The authors discuss especially the implications of this early use of intonation to detect politeness, relating it to other previous research on children’s ability to infer meaning from pitch.

Identification of Focus Information with Contextual and Lexical Cues by Korean Learners of English: Evidence from Reaction Times Analyzed by Linear Mixed-Effect Models

Identification of Focus Information with Contextual and Lexical Cues by Korean Learners of English: Evidence from Reaction Times Analyzed by Linear Mixed-Effect Models

On the relation between lexical and prosodic cues of emotion in English: a preliminary corpus-based analysis

De nombreuses etudes ont montre que la parole emotionnelle est caracterisee par des indices sur plusieurs niveaux linguistiques. Dans cet article, nous analysons la relation entre les indices lexicaux et prosodiques des emotions en anglais dans les enregistrements de livres audios du corpus LibriSpeech. Afin de quantifier les indices lexicaux des emotions, nous avons utilise les outils d’analyse des sentiments disponibles dans la librairie Python NLTK (notamment Vader et SentiWordNet). En revanche, pour la quantification des indices prosodiques des emotions, nous avons extrait plusieurs parametres acoustiques indiquant les variations de la frequence fondamentale, de la qualite de la voix et du rythme. La relation entre indices lexicaux et prosodiques a ete ensuite analysee a travers des modeles lineaires a effets mixtes, qui ont revele des effets faibles mais significatifs. En particulier, seules les variations de F0 sembleraient liees a la valence des phrases et des mots, alors que les variations de F0, du rythme et de la qualite de la voix sembleraient liees au degre de positivite / negativite des phrases et des mots. Cependant, nos modeles n'expliquent qu'une faible part de variance des donnees. Nous estimons donc que la relation entre les indices lexicaux et prosodiques dans les enregistrements de livres audios existe mais est plutot faible. Il est possible que cette observation puisse s’appliquer de maniere plus generale aux cas de conceptualisation ou description des emotions pour lesquelles le locuteur ne se sent pas directement implique. En revanche, il est possible que la relation entre les indices lexicaux et prosodiques s’avere plus etroite dans les cas d’expression des emotions ou le locuteur se sentirait plus directement implique. De toute maniere, nos resultats ne doivent etre consideres que comme preliminaires, etant donne les limites d'une analyse entierement automatique.

Resolving Indeterminacy in Gender Agreement: Comparing Heritage Speakers and L2 Learners of Russian

Cross-linguistically, both heritage language (HL) speakers and second language (L2) learners have been shown to experience difficulty in producing and interpreting linguistic structures characterized by indeterminacy, or lack of an invariable and transparent relationship between meaning and form. This article compares two populations of Russian-English bilinguals on their strategies of resolving ambiguity within the system of grammatical gender in Russian, with a particular focus on indeterminacy in gender agreement with animate nouns. As a result of complex interactions among lexical, morpho-phonological, and discourse-level gender categorization cues, the agreement behavior of animate nouns in Russian is not fully uniform. The results of a scaled acceptability ratings study demonstrate that gender agreement in transparent and non-ambiguous contexts is largely unproblematic for both bilingual groups; however, contexts that require conflict resolution between different types of cues and those characterized by underspecification represent two areas where HL speakers and L2 learners diverge from monolingual Russian-speaking controls. Across all experimental conditions, bilingual speakers demonstrate a higher reliance on morpho-phonological gender categorization cues and assign less weight to lexical and referential factors in gender assignment than monolinguals. The results further show that the two populations of bilinguals are not fully alike with respect to dealing with different types of indeterminacy. In particular, HL speakers exhibit an advantage over L2 learners in conflict resolution; however, both bilingual groups struggle with constructions that give rise to referential ambiguity due to underspecification. These results expand our understanding of the problem of indeterminacy in bilingual acquisition of gender and offer implications for theories of language acquisition and language instruction.