Language Input Research Articles

Building a better lawyer: Experimental evidence that artificial intelligence can increase legal work efficiency

AbstractRapidly improving artificial intelligence (AI) technologies have created opportunities for human–machine cooperation in legal practice. We provide evidence from an experiment with law students (N = 206) on the causal impact of machine assistance on the efficiency of legal task completion in a private law setting with natural language inputs and multidimensional AI outputs. We tested two forms of machine assistance: AI‐generated summaries of legal complaints and AI‐generated text highlighting within those complaints. AI‐generated highlighting reduced task completion time by 30% without any reduction in measured quality indicators compared to no AI assistance. AI‐generated summaries produced no change in performance metrics. AI summaries and AI highlighting together improved efficiency but not as much as AI highlighting alone. Our results show that AI support can dramatically increase the efficiency of legal task completion, but finding the optimal form of AI assistance is a fine‐tuning exercise. Currently, AI‐generated highlighting is not readily available from state‐of‐the‐art, consumer‐facing large language models, but our work suggests that this capability should be prioritized in the development of legal AI products.

Language Dominance in the Input and Use Patterns of Cameroon’s Special Bilingual Students

This paper examines dominance in the language input and use patterns of students of Cameroon’s Special Bilingual Education Programme (SBEP). The participants were 373 bilingual students from 10 schools set in French-speaking Cameroon, who submitted self-ratings via an adapted version of the Bilingual Language Profile questionnaire by Birdsong, Gertken and Amengual (2012). The data were analysed using the subtraction method to determine dominance. The results showed a consistent pattern of OL1-dominance for all the three domains considered under language input (viz. listening, reading, and watch time), but a more contrasted pattern for language use, as the respondents were found to be French-dominant (or slightly OL1-dominant to balanced) in the outer circle of relationships (circle of friends and school environment) and OL1-dominant in the inner circle (self-talk and family circle). The findings of this study suggest that individual language use is more significantly impacted by societal dominance than language input is, most likely because the latter is less constrained by relationships and offers a greater possibility of choice. They equally suggest that there is a directionality in the influence of societal dominance, with a decreasing effect from the public to the private sphere. Lastly, the study underscores the powerful negative impact the lack of official recognition can have on the maintenance of local ethnic languages in a multilingual context.

Utilizing YouTube Videos to Improve Speaking Skills in Seventh-Grade Students at MTsN 2 Kota Bengkulu

This study aimed to explore how YouTube digital videos could enhance students’ speaking skills and to identify the factors that influenced these improvements. Conducted as Classroom Action Research, the study involved 36 seventh-grade students at MTsN 02 Kota Bengkulu. The research employed a mixed-methods approach, gathering both quantitative data through speaking tests and qualitative data through observations, field notes, and interviews. The results revealed that key factors driving this improvement included increased student motivation and active participation in the learning process. Students reported feeling more confident in their speaking abilities, and the authentic language input provided by YouTube videos helped them improve their pronunciation, vocabulary, and fluency. The novelty of this study lay in applying YouTube as a pedagogical tool within an Indonesian junior high school context, specifically focusing on its impact on speaking skills. Furthermore, the study highlighted the importance of combining digital media with interactive classroom activities to create a more engaging and effective learning environment. This research contributed to the growing body of literature on digital media in education by offering practical insights into how YouTube videos could be successfully integrated into language teaching to foster speaking skill development among younger learners. The findings suggested that YouTube digital videos served as an effective medium for improving language proficiency, particularly in EFL (English as a Foreign Language) contexts, where authentic language exposure was often limited. Moreover, the study emphasized the crucial role of motivation and student engagement, offering valuable strategies for educators to implement in their classrooms to enhance student learning outcomes.

What role does the environment play in language development? Exploring the associations among socioeconomic status, parent language input, and language skills in school-aged children with autism.

Language development in children with autism is influenced by proximal (e.g., parent language input) and distal (e.g., socioeconomic status) environmental constructs. Studies have found that "rich and responsive" parent language input supports autistic children's language development, and recent work has reported positive associations between measures of socioeconomic status (SES) and child language skills. However, little is known about how these proximal and distal environmental constructs interact to shape language development in autism. In a sample of 74 autistic school-aged children, the present study investigated the associations among measures of SES, the quantity and quality of language produced by parents and children during home-based dyadic parent-child interactions, and children's expressive and receptive language skills. Results showed that annual household income was positively associated with parent number of total words (NTW), parent number of different words (NDW), and parent mean length of utterance (MLU), while neither parent education level nor annual household income were significantly associated with measures of child language skills. Parent MLU was positively associated with child MLU and child expressive language skills. Findings suggest that annual household income may influence both the quantity and quality of parent language input, and that parent MLU, a qualitative measure of parent language input, may play a particularly important role in shaping autistic children's expressive language development. Future research should study longitudinal associations among SES, parent language input, and child language skills, as identifying environmental predictors of language skills in autism may facilitate the creation of more effective interventions that support language development.

Digital phenotyping of depression during pregnancy using self-report data.

Depression is a common pregnancy complication yet is often under-detected and, subsequently, undertreated. Data collected through mobile health tools may be used to support the identification of depression symptoms in pregnancy. An observational cohort study of 2062 pregnancies collected self-reports of patient history, mood, pregnancy-specific symptoms, and written language using a prenatal support app. These app inputs were used to model depression risk in subsequent 30- and 60-day periods throughout pregnancy. A selective inference lasso modeling approach examined the individual and additive value of each type of patient-reported app input. Depression models ranged in predictive power (AUC value of 0.64-0.83), depending on the type of inputs. The most predictive model included personal history, daily mood, and acute pregnancy-related symptoms (e.g., severe vomiting, cramping). Across models, daily mood was the strongest indicator of depression symptoms in the following month. Models that retained natural language inputs typically improved predictive accuracy and offered insight into the lived context associated with experiencing depression. Our findings are not generalizable beyond a digitally literate patient population that is self-motivated to report data during pregnancy. Simple patient reported data, including sparse language, shared directly via digital tools may support earlier depression symptom identification and a more nuanced understanding of depression context.

Affective Filter Hypothesis and Second Language acquisition: A study on anxiety levels in the dictation part of TEM-4

The Affective Filter Hypothesis is a cornerstone of second language acquisition (SLA) theory. This hypothesis explores the influence of emotional factors on language learning, particularly the role of emotional states such as anxiety, motivation, and self-confidence in the process of language input. This work investigates the applicability of the Affective Filter Hypothesis in the context of the Dictation section of the Test for English Majors, Band 4 (TEM-4). To explore the impact of emotional factors, especially anxiety levels, on performance in the Dictation section of the TEM-4, a mixed-methods approach was employed in this paper. Participants underwent an anxiety level test and a simulated dictation test to establish baseline data. The results indicated that the overall performance trend of the low-anxiety group was higher than that of the high-anxiety group. This outcome suggests that learners with low anxiety level performed better in TEM-4 dictation than learners with high anxiety level.

Investigating EFL English-Majored Students’ Perceptions towards the Use of Technology in Language Teaching

This study investigates EFL English-majored students’ opinions of technology use in language learning at Nam Can Tho University, with an emphasis on both benefits and problems. The study uses a questionnaire survey with 85 participants to evaluate how technology affects students’ engagement, motivation, and overall learning experience. The findings show that students see major benefits from technology, such as increased engagement through multimedia materials, tailored learning experiences via applications, and access to real language input through online tools. Furthermore, students love the opportunity to communicate in real time with native speakers, which improves their language abilities and confidence. In contrast, the research cites a number of hurdles, including difficulty in integrating technology with the curriculum, technical concerns, possible diversions, and the need for more training and support. These findings emphasize the necessity of resolving these issues in order to maximize the usefulness of technology in language instruction, and they offer practical recommendations for educators and legislators.

Developing automaticity in neural speech discrimination in typically developing bilingual Italian-German and monolingual German children.

Many studies have shown that input in more than one language influences children's phonemic development. In this study, we examined the neural processes supporting perception of Voice Onset Time (VOT) in bilingual Italian-German children and their monolingual German peers. While German contrasts short-lag and long-lag, Italian contrasts short-lag and voicing lead. We examined whether bilinguals' phonetic/phonological systems for the two languages develop independently or whether they influence each other, and what role language input plays in the formation of phonetic/phonological categories. Forty five-year-old children (16 monolingual German, 24 bilingual Italian-German) were tested in an oddball design expected to elicit a neural Mismatch Response (MMR). The stimuli were bilabial stop VOT contrasts with the short-lag stop, common to both languages, as the standard. Four deviant VOTs were selected: 92 ms and 36 ms lag for German; 112 ms and 36 ms voicing lead for Italian. Bilingual children's language background was assessed using a caregiver questionnaire. Italian-German bilingual 5-year-old children and German monolingual controls showed similar MMRs to German long-lag and Italian voicing lead VOT, except for the 36 ms long-lag deviant; this acoustically difficult distinction did not elicit a robust negative MMR in the bilingual children. The lack of a difference between the bilinguals and monolinguals for voicing lead suggests that the amount of input in Italian for the bilinguals was not sufficient to lead to an advantage compared to the monolingual German children. Alternatively, the finding could indicate that voicing lead is easier to discriminate than voicing lag.

Listening comprehension in a home language: a case of Russian in Germany.

Listening comprehension is central to language learning, yet it remains the least understood and least researched skill. This statement is still relevant today, as there is insufficient research to explore listening comprehension from the perspective of family-related multilingualism and to consider the complete linguistic repertoire of multilingual speakers. Moreover, with regard to home language, listening comprehension is assumed to be a more developed language competence than reading or writing. Based on the mentioned research, the aim of the present study is to investigate listening comprehension and its influencing factors specifically in German-Russian simultaneous bilinguals aged 13-19 (n = 99) by considering the home- and majority language. The study uses quantitative data collection methods such as linguistic tests in Russian and German for the elicitation in different levels of listening and questionnaires for strategy use and background. The research questions are as follows: What does the language proficiency and input in Russian look like? (1) Regarding listening comprehension in Russian as a home language, are there differences between the speakers within comprehension on different levels, e.g., is sound decoding easier than sentence parsing? (2) If there are differences in Russian as a home language, which linguistic and background variables can correlate with the performance of listening comprehension on its different levels? Concerning the first research question, the results show relevant differences between four different levels of listening comprehension (phoneme, word, sentence, and text level), which strengthened the assumed complexity of listening comprehension in the home language. In addition, the results show different connections between the listening comprehension competence and the input from different family members, as well as exposure to film and television in the home and majority language.

Lexical frequency and target language input in the primary languages classroom

ABSTRACT This paper presents an exploratory study of L2 input in the primary languages classroom. Teacher talk in 33 videorecorded lessons taught to Year 3 French beginners by a specialist teacher of French was analysed, in order to document the extent and nature of L2 vocabulary input. Further analyses compared the lexis of teacher talk in this primary setting with high-frequency wordlists derived from two reference corpora designed to support the teaching of L2 French. Results showed that children were exposed to c 45,000 words (tokens) of French, reflecting a teacher vocabulary of c 650 lemmas. Intentionally taught lexis such as foods or animal names typically received massed practice, not always supplemented by distributed practice. The teachers’ partial but consistent use of French for classroom management contributed to a well-balanced distribution of word classes. There was considerable overlap with both reference wordlists, and divergence could largely be explained by the teacher’s choice of topics, and incidental use of French to personalise and engage childrens’ individual interests, as well as for classroom management purposes. In discussion, the need for careful planning and management of vocabulary instruction is highlighted, as well as the significant potential of L2 classroom management to enrich lexical input.

Vision-and-language navigation based on history-aware cross-modal feature fusion in indoor environment

Vision-and-language navigation (VLN) is a challenging task that requires an agent to navigate an indoor environment using natural language instructions. Traditional VLN employs cross-modal feature fusion, where visual and textual information are combined to guide the agent’s navigation. However, incomplete use of perceptual information, scarcity of domain-specific training data, and diverse image and language inputs result in suboptimal performance. Herein, we propose a cross-modal feature fusion VLN history-aware information, that leverages an agent’s past experiences to make more informed navigation decisions. The regretful model and self-monitoring models are added, and the advantage actor critic(A2C) reinforcement learning algorithm is employed to improve the navigation success rate, reduce action redundancy, and shorten navigation paths. Subsequently, a data augmentation method based on speaker data is introduced to improve the model generalizability. We evaluate the proposed algorithm on the room-to-room (R2R) and room-for-room (R4R) benchmarks, and the experimental results demonstrate that, by comparison, the proposed algorithm outperforms state-of-the-art methods.

The Construction of Personalized Language Learning System under the Guidance of ChatGPT

With the development of artificial intelligence technology, AI-powered chatbots are gradually being applied to language teaching and learning. Among them, ChatGPT, as a generative language intelligence model, plays its unique advantage in promoting personalized language learning. Hence, this article starts by elucidating recent research on the use of ChatGPT in language teaching and learning, giving some evidence about the possibility of personalized language learning under the guidance of ChatGPT. Then advantages associated with its usage in personalized language learning are discussed, including providing personalized language input, cultivating personalized language learning abilities and giving personalized language feedback for language learners. Based on these advantages, this article compares the traditional personalized language learning with the one under the guidance of ChatGPT and try to put forward a three-step personalized language learning system under the guidance of ChatGPT, hoping to improve the effectiveness of language learning and realize the truly personalized language learning in the future.

Are Late Talkers Just Late? Neighborhood Density and Word Frequency Properties of Late Talkers' Spoken Vocabularies.

Typically developing toddlers extract patterns from their input to add words to their spoken lexicons, yet some evidence suggests that late talkers leverage the statistical regularities of the ambient language differently than do peers. Using the extended statistical learning account, we sought to compare lexical-level statistical features of spoken vocabularies between late talkers and two typically developing comparison groups. MacArthur-Bates Communicative Developmental Inventories American English Words and Sentences (N = 1,636) were extracted from Wordbank, a database of CDIs. Inventories were divided into three groups: (a) a late talker group (n = 202); (b) a typically developing age-matched group (n = 1,238); and (c) a younger, typically developing group (n = 196) matched to the late talkers on expressive language. Neighborhood density and word frequency were calculated for each word produced by each participant and standardized to z scores. Mixed-effects models were used to evaluate group differences. The late talker and younger, language-matched groups' spoken vocabularies consist, on standardized average, of words from denser phonological neighborhoods and words higher in frequency of occurrence in parent-child speech, compared to older, typically developing toddlers. These findings provide support for the extended statistical learning account. Late talkers appear to generally be extracting and using similar patterns from their language input as do younger toddlers with similar levels of expressive vocabulary. This suggests that late talkers may be following a delayed, not deviant, trajectory of expressive language growth.

Accelerated Data Engine: A faster dataset construction workflow for computer vision applications in commercial livestock farms

Large-scale, high-quality dataset was the foundation of developing advanced artificial intelligence applications. However, creating such a benchmark dataset in a professional field, such as precision management of animals, was always a challenge because of the costly and labor-intensive process of annotation and review. This study introduced a novel workflow named Accelerated Data Engine (ADE), designed to efficiently produce representative and high-quality computer vision datasets from raw animal surveillance footage. By incorporating referring and grounding models (R&G models) as auto-annotators, along with a distillation mechanism for dataset-auditors, ADE significantly speeded up the dataset construction process. The new workflow received natural language inputs as referrals to identify animal instances, delineated their body shapes, and then refined the auto-annotated data through a selection process. To demonstrate the efficacy of ADE, three 30-minute surveillance video samples featuring pigs, sheep, and cattle were discussed in this study. The results indicated the R&G models effectively annotated animals across various farms, while distillation mechanisms could identify various detection errors, balance the data representations, refine annotations, and verify the data quality. Two high-quality cattle datasets (6.5 k and 486 frames), including 26 k and 2.5 k cattle instances, were generated through the ADE workflow from 24-hour surveillance videos on a commercial cattle farm and made publicly available. The proposed dataset has achievable performance between 74.6 %∼84.1 %. The ADE workflow saved 78.4 % of manual work compared to the traditional dataset construction workflow (approximately 141 h). This pioneering approach empowered the fast creation of benchmark animal datasets and would enhance computer vision applications in the livestock production industry in the future.

Performance of ChatGPT on Nursing Licensure Examinations in the United States and China: Cross-Sectional Study.

The creation of large language models (LLMs) such as ChatGPT is an important step in the development of artificial intelligence, which shows great potential in medical education due to its powerful language understanding and generative capabilities. The purpose of this study was to quantitatively evaluate and comprehensively analyze ChatGPT's performance in handling questions for the National Nursing Licensure Examination (NNLE) in China and the United States, including the National Council Licensure Examination for Registered Nurses (NCLEX-RN) and the NNLE. This study aims to examine how well LLMs respond to the NCLEX-RN and the NNLE multiple-choice questions (MCQs) in various language inputs. To evaluate whether LLMs can be used as multilingual learning assistance for nursing, and to assess whether they possess a repository of professional knowledge applicable to clinical nursing practice. First, we compiled 150 NCLEX-RN Practical MCQs, 240 NNLE Theoretical MCQs, and 240 NNLE Practical MCQs. Then, the translation function of ChatGPT 3.5 was used to translate NCLEX-RN questions from English to Chinese and NNLE questions from Chinese to English. Finally, the original version and the translated version of the MCQs were inputted into ChatGPT 4.0, ChatGPT 3.5, and Google Bard. Different LLMs were compared according to the accuracy rate, and the differences between different language inputs were compared. The accuracy rates of ChatGPT 4.0 for NCLEX-RN practical questions and Chinese-translated NCLEX-RN practical questions were 88.7% (133/150) and 79.3% (119/150), respectively. Despite the statistical significance of the difference (P=.03), the correct rate was generally satisfactory. Around 71.9% (169/235) of NNLE Theoretical MCQs and 69.1% (161/233) of NNLE Practical MCQs were correctly answered by ChatGPT 4.0. The accuracy of ChatGPT 4.0 in processing NNLE Theoretical MCQs and NNLE Practical MCQs translated into English was 71.5% (168/235; P=.92) and 67.8% (158/233; P=.77), respectively, and there was no statistically significant difference between the results of text input in different languages. ChatGPT 3.5 (NCLEX-RN P=.003, NNLE Theoretical P<.001, NNLE Practical P=.12) and Google Bard (NCLEX-RN P<.001, NNLE Theoretical P<.001, NNLE Practical P<.001) had lower accuracy rates for nursing-related MCQs than ChatGPT 4.0 in English input. English accuracy was higher when compared with ChatGPT 3.5's Chinese input, and the difference was statistically significant (NCLEX-RN P=.02, NNLE Practical P=.02). Whether submitted in Chinese or English, the MCQs from the NCLEX-RN and NNLE demonstrated that ChatGPT 4.0 had the highest number of unique correct responses and the lowest number of unique incorrect responses among the 3 LLMs. This study, focusing on 618 nursing MCQs including NCLEX-RN and NNLE exams, found that ChatGPT 4.0 outperformed ChatGPT 3.5 and Google Bard in accuracy. It excelled in processing English and Chinese inputs, underscoring its potential as a valuable tool in nursing education and clinical decision-making.

A Corpus-Based Multifactorial Study of Help/help to Alternation in Learners’ Language: From the Perspective of Probabilistic Grammar

This corpus-based multifactorial study delves deeper into the well-known alternation between bare and full infinitive complements, specifically regarding the help concordances. It extends the line of research to learners’ language productions with a focus on comparing and contrasting the probabilistic grammatical knowledge reflected in the help/ help to choices, which are constrained by a complex network of factors. Regression modeling indicates that constraints such as the horror aequi and certain morphological forms of the verb help are comparatively stable among groups of Chinese English language learners and native English speakers. Therefore, the principle of cognitive complexity and the avoidance strategy fueling infinitive variations are further tested. While the object typology differs in effect strength, it also prompts consideration of cultural influences on the part of Chinese English learners. In addition, two different patterns of interaction effects are explored in each dataset, which implies that learners’ language choice is usage-based, determined by an unconscious assessment of multiple clues that are learned from language input and shaped by language experiences.

Transforming Creative Process: A Systematic Literature Review of Discourse on AI Image Generators

ABSTRACTAI image generator, also known as Text‐to‐image generation (TTIG) technology, allows users to generate images based on natural language input. With advancements in AI, debates related to its application in the creative field have flourished. A systematic literature review was conducted to examine discourse that addresses the perception of people in the creative industry towards AI image generators and the impact of AI image generators on these individuals. The analysis of discourse revealed four major categories: 1) Evaluation of product or outcome of AI; 2) Impact of AI on practitioners/artists; 3) Impact of AI on creative industry and society; 4) Capabilities of AI technology. These findings provide insights into the areas of interest within AI image generator discussions on creative activity.

The unique contribution of uncertainty reduction during naturalistic language comprehension

Language comprehension is an incremental process with prediction. Delineating various mental states during such a process is critical to understanding the relationship between human cognition and the properties of language. Entropy reduction, which indicates the dynamic decrease of uncertainty as language input unfolds, has been recognized as effective in predicting neural responses during comprehension. According to the entropy reduction hypothesis (Hale, 2006), entropy reduction is related to the processing difficulty of a word, the effect of which may overlap with other well-documented information-theoretical metrics such as surprisal or next-word entropy. However, the processing difficulty was often confused with the information conveyed by a word, especially lacking neural differentiation. We propose that entropy reduction represents the cognitive neural process of information gain that can be dissociated from processing difficulty. This study characterized various information-theoretical metrics using GPT-2 and identified the unique effects of entropy reduction in predicting fMRI time series acquired during language comprehension. In addition to the effects of surprisal and entropy, entropy reduction was associated with activations in the left inferior frontal gyrus, bilateral ventromedial prefrontal cortex, insula, thalamus, basal ganglia, and middle cingulate cortex. The reduction of uncertainty, rather than its fluctuation, proved to be an effective factor in modeling neural responses. The neural substrates underlying the reduction in uncertainty might imply the brain’s desire for information regardless of processing difficulty.

Distribution of words across the first years of life: A longitudinal analysis of everyday language input to three English-learning infants.

Many in-lab studies have demonstrated that the distribution of word learning moments affects the strength and quality of word representations. How are words distributed in speech to children in their daily lives, and how is distribution related to other input characteristics? The present study analyzes transcripts of language input to English-learning infants from three longitudinal, naturalistic corpora captured between 6 and 39months of age. To describe how word frequency varies across time, we calculated dispersion scores for all word types for each child. Dispersion quantifies the deviation of observed frequencies in each recording session from expected (uniform across sessions) word frequency, providing a measure of how evenly word utterances were spread across sessions. Dispersion is strongly correlated with frequency and moderately correlated with concreteness across all corpora, such that high frequency and low concreteness words are more evenly dispersed. Correlations with measures of age of acquisition (AoA) varied across corpora, and dispersion did not reliably predict AoA above and beyond frequency and concreteness. The contradiction between the current results and results from in-lab experiments is discussed. This study provides a foundation to explore how word learning unfolds across time and contexts in the real world.

ExTraCT - Explainable trajectory corrections for language-based human-robot interaction using textual feature descriptions.

In human-robot interaction (HRI), understanding human intent is crucial for robots to perform tasks that align with user preferences. Traditional methods that aim to modify robot trajectories based on language corrections often require extensive training to generalize across diverse objects, initial trajectories, and scenarios. This work presents ExTraCT, a modular framework designed to modify robot trajectories (and behaviour) using natural language input. Unlike traditional end-to-end learning approaches, ExTraCT separates language understanding from trajectory modification, allowing robots to adapt language corrections to new tasks-including those with complex motions like scooping-as well as various initial trajectories and object configurations without additional end-to-end training. ExTraCT leverages Large Language Models (LLMs) to semantically match language corrections to predefined trajectory modification functions, allowing the robot to make necessary adjustments to its path. This modular approach overcomes the limitations of pre-trained datasets and offers versatility across various applications. Comprehensive user studies conducted in simulation and with a physical robot arm demonstrated that ExTraCT's trajectory corrections are more accurate and preferred by users in 80% of cases compared to the baseline. ExTraCT offers a more explainable approach to understanding language corrections, which could facilitate learning human preferences. We also demonstrated the adaptability and effectiveness of ExTraCT in a complex scenarios like assistive feeding, presenting it as a versatile solution across various HRI applications.