Computer-assisted Language Learning Research Articles

The Proteasome-Family-Members-Based Prognostic Model Improves the Risk Classification for Adult Acute Myeloid Leukemia.

Background: The accumulation of diverse molecular and cytogenetic variations contributes to the heterogeneity of acute myeloid leukemia (AML), a cluster of hematologic malignancies that necessitates enhanced risk evaluation for prognostic prediction and therapeutic guidance. The ubiquitin-proteasome system plays a crucial role in AML; however, the specific contributions of 49 core proteasome family members (PSMs) in this context remain largely unexplored. Methods: The expression and survival significance of 49 PSMs in AML were evaluated using the data from BeatAML2.0, TCGA, and the GEO database, mainly through the K-M plots, differential genes enrichment analysis, and candidate compounds screening via R language and statistical software. Results: we employed LASSO and Cox regression analyses and developed a model comprising three PSMs (PSMB8, PSMG1, and PSMG4) aimed at predicting OS in adult AML patients, utilizing expression profiles from the BeatAML2.0 training datasets. Patients with higher risk scores were predominantly found in the AML-M2 subtype, exhibited poorer ELN stratification, showed no complete remission following induction therapies, and had a higher mortality status. Consistently, significantly worse OS was observed in high-risk patients across both the training and three validation datasets, underscoring the robust predictive capability of the three-PSMs model for AML outcomes. This model elucidated the distinct genetic abnormalities landscape between high- and low-risk groups and enhanced the ELN risk stratification system. Ultimately, the three-PSMs risk score captured AML-specific gene expression signatures, providing a molecular basis for selecting potential therapeutic agents. Conclusions: In summary, these findings manifested the significant potential of the PSM model for predicting AML survival and informed treatment strategies.

Identification of Salt-Stress-Responding Genes by Weighted Gene Correlation Network Analysis and Association Analysis in Wheat Leaves.

The leaf is not only the main site of photosynthesis, but also an important organ reflecting plant salt tolerance. Discovery of salt-stress-responding genes in the leaf is of great significance for the molecular improvement of salt tolerance in wheat varieties. In this study, transcriptome sequencing was conducted on the leaves of salt-tolerant wheat germplasm CH7034 seedlings at 0, 1, 6, 24, and 48 h after NaCl treatment. Based on weighted gene correlation network analysis of differentially expressed genes (DEGs) under salt stress, 12 co-expression modules were obtained, of which, 9 modules containing 4029 DEGs were related to the salt stress time-course. These DEGs were submitted to the Wheat Union database, and a total of 904,588 SNPs were retrieved from 114 wheat germplasms, distributed on 21 wheat chromosomes. Using the R language package and GAPIT program, association analysis was performed between 904,588 SNPs and leaf salt injury index of 114 wheat germplasms. The results showed that 30 single nucleotide polymorphisms (SNPs) from 15 DEGs were associated with salt tolerance. Then, nine candidate genes, including four genes (TaBAM, TaPGDH, TaGluTR, and TaAAP) encoding enzymes as well as five genes (TaB12D, TaS40, TaPPR, TaJAZ, and TaWRKY) encoding functional proteins, were identified by converting salt tolerance-related SNPs into Kompetitive Allele-Specifc PCR (KASP) markers for validation. Finally, interaction network prediction was performed on TaBAM and TaAAP, both belonging to the Turquoise module. Our results will contribute to a further understanding of the salt stress response mechanism in plant leaves and provide candidate genes and molecular markers for improving salt-tolerant wheat varieties.

Integration of Computer-Assisted Language Instruction in English Language Teaching of Oral Skills among Secondary School Learners

Integration of Computer-Assisted Language Instruction in English Language Teaching of Oral Skills among Secondary School Learners

Vegetation Phenology Changes and Recovery after an Extreme Rainfall Event: A Case Study in Henan Province, China

Extreme rainfall can severely affect all vegetation types, significantly impacting crop yield and quality. This study aimed to assess the response and recovery of vegetation phenology to an extreme rainfall event (with total weekly rainfall exceeding 500 mm in several cities) in Henan Province, China, in 2021. The analysis utilized multi-sourced data, including remote sensing reflectance, meteorological, and crop yield data. First, the Normalized Difference Vegetation Index (NDVI) time series was calculated from reflectance data on the Google Earth Engine (GEE) platform. Next, the ‘phenofit’ R language package was used to extract the phenology parameters—the start of the growing season (SOS) and the end of the growing season (EOS). Finally, the Statistical Package for the Social Sciences (SPSS, v.26.0.0.0) software was used for Duncan’s analysis, and Matrix Laboratory (MATLAB, v.R2022b) software was used to analyze the effects of rainfall on land surface phenology (LSP) and crop yield. The results showed the following. (1) The extreme rainfall event’s impact on phenology manifested directly as a delay in EOS in the year of the event. In 2021, the EOS of the second growing season was delayed by 4.97 days for cropland, 15.54 days for forest, 13.06 days for grassland, and 12.49 days for shrubland. (2) Resistance was weak in 2021, but recovery reached in most areas by 2022 and slowed in 2023. (3) In each year, SOS was predominantly negatively correlated with total rainfall in July (64% of cropland area in the first growing season, 53% of grassland area, and 71% of shrubland area). In contrast, the EOS was predominantly positively correlated with rainfall (51% and 54% area of cropland in the first and second growing season, respectively, and 76% of shrubland area); however, crop yields were mainly negatively correlated with rainfall (71% for corn, 60% for beans) and decreased during the year of the event, with negative correlation coefficients between rainfall and yield (−0.02 for corn, −0.25 for beans). This work highlights the sensitivity of crops to extreme rainfall and underscores the need for further research on their long-term recovery.

Technology integration into Chinese as a foreign language learning in higher education: An integrated bibliometric analysis and systematic review (2000–2024)

As Chinese as a Foreign Language (CFL) becomes increasingly popular worldwide, it is also influenced by modern educational technology. This article seeks to comprehensively review the current research status of technology integration into CFL learning in higher education. Using the Web of Science as the primary database and employing VOSviewer and CitNetExplorer to visualize literature, we conducted bibliometric analysis and systematically reviewed 64 included articles published between 2000 and 2024. Visualization analysis shows popular issues, such as technology acceptance, technology-supported collaborative learning, Technological, Pedagogical, and Content Knowledge (TPACK), self-efficacy, and gender stereotypes. Educational technologies possess characteristics highly adaptable to CFL classrooms in higher education, with mobile- and computer-assisted language learning dominating the field. Regarding CFL skills, most studies focused on technology’s role in enhancing students’ speaking and communication skills. Overall, our research contributes valuable insights by summarizing the effectiveness of technology application in CFL higher education, paving the way for future studies and practices to further explore how technology can be maximally integrated into classrooms to enhance CFL students’ proficiency and learning experiences.

Online language teaching in times of change: a CALL to action for language teachers and educators

Abstract Change, whether slowly developing over decades or rapid change caused by worldwide crises, affects education and educators profoundly. This article argues for a change in language teacher education embracing the changing world of online communication, and emphasising the epistemological awareness language teachers need to bring to the online classroom. Based on Computer-Assisted Language Learning (CALL) research and best practice examples from across the world, the paper outlines past changes, contextualises online language teaching in the wider area of online communication and describes progress in practice, theory and epistemological understanding of online teaching. It culminates in a series of evidence-based suggestions for future training and education of language teachers that will integrate adequate pedagogy and an enhanced epistemological understanding.

An interactive R-based custom quantification program for semi-quantitative analysis of triacylglycerols in bovine milk.

The R programming language, RStudio, and open-source software solutions for analysis of liquid chromatography-mass spectrometry (LC-MS) data have been used with user-written R-based custom quantification programs (CQP) for semi-quantification of triacylglycerols (TAGs) in bovine milk lipid extracts. Using the peak-finding capabilities of the package "xcms" in RStudio, peaks were integrated, and retention times aligned, normalized, and then used for semi-quantitative analysis of a custom set of four extraction internal standards (EISs) and 29 TAG regioisomers using the choice of four analytical internal standards (AISs). Alternating stereospecific numbering (sn) 1,3 TAG regioisomers (standards 1, 3, and 5 of six calibration standards) and sn-1,2 TAG regioisomers (standards 2, 4, and 6 of six standards) were used to make a set of six calibration standards, which were used for quantification using a linear fit model, polynomial fit model, power fit model, level-bracketed linear fit, replicate-bracketed polynomial fit, replicate-bracketed power fit, and replicate- and level-bracketed linear fit and response factors. For example, the linear fit for EIS1 gave an unacceptable coefficient of determination (CoD), r2 = 0.9616, whereas the polynomial fit gave r2 = 0.9908 and the power fit gave r2 = 0.9928, while the double-bracketed linear fit gave CoDs of r2 = 0.9960, 0.9848, and 0.9781 for the three brackets, yet gave the least % difference to known calibration concentrations. For unparalleled transparency, the CQP produced webpages that allowed every step in the data processing and quantification sequence to be verified and reproduced, and contained interactive figures. The data are publicly available using a digital object identifier (DOI). The R code can be downloaded and used with the downloadable data to reproduce the results, to modify the code and further customize the results, or to copy and paste and adapt the code to other quantification applications.

Prognostic value of PEA3 subfamily gene expression in cholangiocarcinoma.

Cholangiocarcinoma (CCA) is a lethal malignancy with limited treatment options and poor prognosis. The PEA3 subfamily of E26 transformation specific genes: ETV1, ETV4, and ETV5 are known to play significant roles in various cancers by influencing cell proliferation, invasion, and metastasis. To analyze PEA3 subfamily gene expression levels in CCA and their correlation with clinical parameters to determine their prognostic value for CCA. The expression levels of PEA3 subfamily genes in pan-cancer and CCA data in the cancer genome atlas and genotype-tissue expression project databases were analyzed with R language software. Survival curve and receiver operating characteristic analyses were performed using the SurvMiner, Survival, and Procr language packages. The gene expression profiling interactive analysis 2.0 database was used to analyze the expression levels of PEA3 subfamily genes in different subtypes and stages of CCA. Web Gestalt was used to perform the gene ontology/ Kyoto encyclopedia of genes and genomes (GO/KEGG) analysis, and STRING database analysis was used to determine the genes and proteins related to PEA3 subfamily genes. ETV1, ETV4, and ETV5 expression levels were significantly increased in CCA. There were significant differences in ETV1, ETV4, and ETV5 expression levels among the different subtypes of CCA, and predictive analysis revealed that only high ETV1 and ETV4 expression levels were significantly associated with shorter overall survival in patients with CCA. GO/KEGG analysis revealed that PEA3 subfamily genes were closely related to transcriptional misregulation in cancer. In vitro and in vivo experiments revealed that PEA3 silencing inhibited the invasion and metastasis of CCA cells. The expression level of ETV4 may be a predictive biomarker of survival in patients with CCA.

Factors impeding implementing CALL-based instruction: A mixed-methods study

As an innovative approach to language pedagogy, Computer-Assisted Language Learning (CALL) research has been gaining traction in recent decades. However, the novel incorporation of technology in the classroom context presents some unique challenges unprecedented in conventional language learning contexts. The present study is an attempt at identifying the factors prohibiting the successful implementation of CALL, with particular attention on the context of Iran. To this end, an exploratory mixed-methods sequential design was adopted. Fifteen Iranian EFL instructors were interviewed and the data were analyzed through thematic analysis. Built upon the findings of the interviews, a questionnaire was developed to assess the generalizability of the interview findings. The 33-item questionnaire was developed and distributed online among 200 EFL teachers. It was revealed that certain factors, including EFL teachers and learners' knowledge of computers, cultural and religious background, institutional support policy, and human factors such as learners' learning style and age could hinder the successful incorporation of CALL in language learning contexts. Principal Component Analysis (PCA) confirmed the generalizability of the identified prohibitive factors. The findings are discussed under the light of Rogers’ (2003) Diffusion of Innovation Theory (DIT) to further contextualize the implications of the study for macro- and micro-level policymakers, curriculum developers, and language teachers.

Critical Interculturality and CALL in ELT: A Necessary Approach in Colombian Contemporary Education

The most recent events in contemporary history have accelerated not only the evolution of technology but also multiple fields of knowledge which include models of teaching English as a Foreign Language (EFL). However, Colombian education has resisted the changes of the XXI century due to the generational clash between “digital natives” and “digital immigrants” (Prensky, 2001). Furthermore, EFL in Colombian public schools has been faithful to Western models that elude local contexts and contribute little to reflection, development of proposals, and students’ agency regarding our society’s day-to-day reality. Such actions, as proposed by Catherine Walsh, fall within the realm of what Critical Interculturality (CI) promotes (Walsh, 2010). Coupled with this, the integration of Computer Assisted Language Learning (CALL) to develop this approach has been limited since its usage has mainly been focused on improving linguistic skills. In this article I will present some theoretical foundations and empirical studies that unveil the importance of developing more studies that facilitate the integration of CI and CALL tools in EFL at the public secondary school in Colombia.

AI-Enhanced Cross-Cultural Competence in STEM Education

Cross-cultural competence (CCC) is leaders' most critical skill set in science, technology, engineering, and math (STEM) education because it can help them effectively manage cultural diversity. The research provides insights into the potential of AI to foster more inclusive and culturally aware educational environments for STEM students using a theory that combines Constructivism and Cultural-Historical Activity Theory (CHAT). The study discusses two research questions: (1) How can AI-driven tools be used in STEM education to enhance cross-cultural sensitivity and communication? (2) What is the impact of AI-facilitated reflective practices on students' self-awareness of cultural biases? This research shows how AI-based language translators and cultural simulations enable learners to adapt to different cultures. According to Constructivism, learning occurs when learners actively construct their knowledge from experiences towards cultural sensitivity, understanding, and respecting other cultures while reducing stereotypes. CHAT, on its part, views AI as aligning with AI as one of the artifacts or tools that are at the disposal of human beings in order to facilitate social interaction; thus, it has the potential to enhance intercultural communication where verbal cues are not only linguistic but also non-verbal. A total of 228 respondents were reached through surveys, resulting in preliminary results showing improved CCC due to accommodative devices such as culturally-specific software simulations or language translators employing AI models. This study adds value to the existing literature by presenting empirical evidence of how AI can foster CCC in STEM education. It highlights the imperative of AI's theoretical and practical applications in educational curricula.

Analyzing the Impact of CALL Tools on English Learners' Writing Skills: A Comparative Study of Errors Correction

The study seeks to compare the effectiveness of Computer-Assisted Language Learning Tools (ChatGPT, Grammarly, and Google Translate) in correcting the occurrences of common errors in English as a second/foreign language (ESL/EFL) learners’ writing. An experimental design offered instructions for Class A with ChatGPT, Class B with Grammarly, Class C with Google Translate, and Class D was the control group with no exposure to any of these tools. Data were collected from texts created by the students based on pictures, both before and after the intervention. It was found that ESL/EFL learners frequently made eight distinct types of errors in their writing: spelling, punctuation, capitalization, possessive words, verb forms, subject-verb agreement, articles, and prepositions. The interventions across different classes showed that ChatGPT in Class A significantly corrected spelling, verb form, and subject-verb agreement-related errors, Grammarly in Class B excelled in correcting prepositional errors, and Google Translate in Class C effectively addressed errors in article use, capitalization, and possessives. In contrast, the control group in Class D surpassed the others in enhancing punctuation. In Class A, ChatGPT, and in Class B, Google Translate displayed enhanced effectiveness in rectifying common writing errors of ESL learners compared to Grammarly in Class B and the control group. The findings suggest that employing targeted digital tools for distinct grammatical challenges can substantially enhance learning outcomes in a language learning context.

Raspberry Pi 3 Based Cloud Supported Temperature Controlled Spectrophotometer Design

Spectrophotometers are a kind of photometers that are used to find the amount of substance in the solution. Spectrophotometers detect the amount of radiation that is not absorbed by the solution in the sample chamber and gives information about the amount of substance in the sample. Spectrophotometers are used in many different fields such as chemistry, medicine, geology, environmental engineering and agriculture. Samples that will directly affect people, such as wastewater, drinking water, pharmaceuticals, are analyzed with high accuracy and possible hazards are avoided. In this study, a device that can be used in the detection of Salmonella bacteria is modeled. In addition, the fact that the heater unit of the device is on the test mechanism eliminates the need for heating in the external environment and the test data can be saved in the cloud environment over the internet connection, which is the novelty of this study. A spectrophotometer device was developed using Raspberry Pi 3 and Python software language. A 470 nm optical filter was used in the device. In addition to the features of standard spectrophotometers, the heating feature that can keep the sample temperature at the desired value has been added. This feature has been developed using the on-off logic method. Linear regression of optical measurements, R2 value was calculated as 0.9982 in density tests between 0.1 gram and 0.2 gram. The device was developed with a Raspberry Pi 3 minicomputer, it does not need an external computer for use. Thanks to the internet connection ability of Raspberry Pi 3, the analysis results are transferred to the cloud environment, allowing the user to see the results even when they are not at the device.

Rethinking AI: bias in speech-recognition chatbots for ELT

Abstract Artificial intelligence (AI) technologies, particularly chatbots with speech-recognition, are gaining attention as tools for ELT. However, this frontline development in contemporary ELT seems to stand in stark contrast to the multilingual effort, another innovative trend, as chatbots’ speech recognition capabilities are primarily attuned to native English speaker pronunciation. Addressing this concern, this study adopted a Global Englishes (GE) perspective and examined students’ attitudes toward English after their use of a speech-recognition chatbot. Text-mining analysis of reflection papers from 38 students and thematic analysis of interviews with seven students revealed that the chatbot hindered the incorporation of diverse English use into ELT and students expressed a conception of English not as a language in flux but rather as a subject with specific content to be mastered. The findings are discussed concerning GE, AI ethics, and computer-assisted language learning, and implications for future chatbot use in ELT are provided.

Research hotspots and future trends in lipid metabolism in chronic kidney disease: a bibliometric and visualization analysis from 2004 to 2023.

Disorders of lipid metabolism play a key role in the initiation and progression of chronic kidney disease (CKD). Recently, research on lipid metabolism in CKD has rapidly increased worldwide. However, comprehensive bibliometric analyses in this field are lacking. Therefore, this study aimed to evaluate publications in the field of lipid metabolism in CKD over the past 20years based on bibliometric analysis methods to understand the important achievements, popular research topics, and emerging thematic trends. Literature on lipid metabolism in CKD, published between 2004 and 2023, was retrieved from the Web of Science Core Collection. The VOSviewer (v.1.6.19), CiteSpace (v.6.3 R1), R language (v.4.3.2), and Bibliometrix (v.4.1.4) packages (https://www.bibliometrix.org) were used for the bibliometric analysis and visualization. Annual output, author, country, institution, journal, cited literature, co-cited literature, and keywords were also included. The citation frequency and H-index were used to evaluate quality and influence. In total, 1,285 publications in the field of lipid metabolism in CKD were identified in this study. A total of 7,615 authors from 1,885 institutions in 69 countries and regions published articles in 466 journals. Among them, China was the most productive (368 articles), and the United States had the most citations (17,880 times) and the highest H-index (75). Vaziri Nosratola D, Levi Moshe, Fornoni Alessia, Zhao Yingyong, and Merscher Sandra emerged as core authors. Levi Moshe (2,247 times) and Vaziri Nosratola D (1,969 times) were also authors of the top two most cited publications. The International Journal of Molecular Sciences and Kidney International are the most published and cited journals in this field, respectively. Cardiovascular disease (CVD) and diabetic kidney disease (DKD) have attracted significant attention in the field of lipid metabolism. Oxidative stress, inflammation, insulin resistance, autophagy, and cell death are the key research topics in this field. Through bibliometric analysis, the current status and global trends in lipid metabolism in CKD were demonstrated. CVD and DKD are closely associated with the lipid metabolism of patients with CKD. Future studies should focus on effective CKD treatments using lipid-lowering targets.

Effects of teachers’ rapport-building strategies on EFL learners’ cognitive load and computer-assisted language learning motivation

Effects of teachers’ rapport-building strategies on EFL learners’ cognitive load and computer-assisted language learning motivation

A Glance at Computer-Assisted Language Learning (CALL): A Bibliometric Analysis

Computer-assisted language Learning (CALL) is a learning method that uses computers as auxiliary media to facilitate language learning. This study aims to provide a comprehensive overview of the development of research in Computer-Assisted Language Learning (CALL) in the last five years based on the Scopus database. This quantitative study uses bibliometric analysis with the search keyword Computer-Assisted Language Learning (CALL). The results of this study highlight a significant increase in the number of articles about Computer-Assisted Language Learning (CALL) indexed by Scopus every year, starting from 142 articles in 2018 and 2023, which increased to 251. The type of publication is dominated by journal articles, followed by conference papers and reviews, and the rest in the form of books or book chapters. The largest source of publications comes from Lecture Notes in Computer Science (including the subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) with 29 publications. Our comprehensive analysis reveals a significant downward trend in citations from 2018 to 2023. Furthermore, based on the visualization using VOSviewer shows the use of AI technology in language teaching and the impact of digital media on language learning. The results of this study show that it is essential to continue to review and evaluate the factors influencing the dissemination and acceptance of CALL research. Moreover, it is crucial to reinforce the continuity between technological innovation and practical practice in various areas of life, thereby highlighting the real-world impact of CALL research.

Novel biomarkers associated with oxidative stress and immune infiltration in intervertebral disc degeneration based on bioinformatics approaches

BackgroundThe etiology of intervertebral disc degeneration (IVDD), a prevalent degenerative disease in the elderly, remains to be fully elucidated. The objective of this study was to identify immune infiltration and oxidative stress (OS) biomarkers in IVDD, aiming to provide further insights into the intricate pathogenesis of IVDD. MethodsThe Gene Expression microarrays were obtained from the Gene Expression Omnibus (GEO) database. We conducted enrichment analysis of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) terms. Subsequently, the R language packages CIBERSORT, MCPcounter, and WGCNA were employed to compare immune infiltration levels between IVDD samples and control samples. A protein-protein interaction (PPI) network was constructed using the Search Tools for the Retrieval of Interacting Genes (STRING) database to identify significant gene clusters. To identify hub genes, we employed Cytoscape's Molecular Complex Detection (MCODE) plug-in. The mRNA levels of hub genes in the cell model were validated by qPCR, while Western blotting was used to validate their protein levels. ResultsThe GSE70362 dataset from the GEO database identified a total of 1799 genes that were differentially expressed. Among these, 43 genes were found to be differentially expressed and also associated with OS. The differentially expressed genes associated with OS and the immune-related module genes identified through WGCNA were further intersected, resulting in the identification of 10 key genes that were differentially expressed and played crucial roles in both immune response and OS. Subsequently, we validated four diagnostic markers (PPIA, MAP3K5, PXN, and JAK2) using the GSE122429 external dataset. In a cellular model of OS in NP cells, we have identified the upregulation of PPIA and PXN genes, which could serve as novel markers for IVDD. ConclusionThe study successfully identified and validated differentially expressed genes associated with oxidative stress and immune infiltration in IVDD samples compared to normal ones. Notably, the newly discovered biomarkers PPIA and PXN have not been previously reported in IVDD-related research.

Computer-assisted DDL vs. Mobile-assisted DDL: Exploring EFL Learners’ Preferences and Experiences with Two Corpus Interfaces

Data-driven learning (DDL) has traditionally been studied via standard corpus interfaces, such as COCA, Sketch Engine, IntelliText, and other similar corpus tools. In addition, DDL has traditionally been seen as part of the computer-assisted language learning (CALL) approach, primarily because access to corpora and corpus interfaces has historically been limited to desktop computers. However, with the emergence of several simplified and mobile-friendly web corpus interfaces and with mobile assisted language learning (MALL) gaining momentum in language education, along with its potential in DDL, mobile-assisted DDL has become an alternative learning technique. This paper reports on a study carried out with a group of undergraduate EFL students studying at International Balkan University (IBU). In this study, the participants were given linguistic tasks to complete using new web corpus interfaces as reference tools, namely, CorpusMate and Netspeak. The first corpus interface was exploited via the traditional CALL approach, whereas the latter was utilized with mobile technology. The results of the study revealed positive learner perceptions and experiences with the newly introduced tools, while they also displayed enthusiasm and motivation to engage in DDL within the context of MALL. The present study provides valuable insights into students’ perspectives and experiences with state-of-the-art corpus tools and mobile technology, which can help shape future paths for research in DDL and MALL.

Research on the impact of streamers’ linguistic emotional valence on live streaming performance in live streaming shopping environments

As an emerging retail model, live streaming shopping involves real-time interactions between streamers and viewers, and among the viewers themselves. In live streaming interactions, the comments posted by the viewers reflect their emotional states, which may have been influenced by the streamers’ emotional expression. This study utilized the S-O-R model and emotional contagion theory as its theoretical foundation to investigate how the emotional valence of streamer speech and real-time comment influence live streaming performance (sales growth/fans growth) and explored the moderating effect of the live streaming atmosphere cues. With Python software and natural language processing technology of Baidu AI platform, sentiment analysis was conducted on 124,085 streamer speech sentences and 24,3767 viewer real-time comments. Our findings reveal that both positive and negative emotional valence of streamer speech positively affect live streaming performance; in addition to the negative impact of neutral emotional valence on sales growth and of negative emotional valence on fans growth, real-time comment emotional valence has a positive correlation with live streaming performance; finally, the study examined the moderating effect of atmosphere cues and demonstrated the complex dynamics of atmospheric cues affecting emotional contagion and sales in live streaming. The findings provide a scientific reference for live marketing managers to optimize the emotional expression of streamer speech and their live streaming operation mode.