Search

Skilled orators and their public speaking skills have been held in high esteem since antiquity. Artificial intelligence-powered presentation platforms (AI-PPPs) are an emerging educational technology that provides a virtual venue and AI audience for users to practice a presentation or pitch and receive feedback to improve their public speaking skills. AI-PPPs allow users to customize their context (e.g., audience type, setting, type of presentation) and make their presentation to a virtual audience comprised of avatars. AI-PPPs use algorithms to interpret the user's public speaking skills and provide an analysis of their performance, often including metrics on pitch, eye contact, pace, volume, distractors, pauses, etc. This research analyzes one such AI-PPP to determine its effectiveness utilizing the text complexity framework. This mixed methods case study reports participants' AI-PPP-assigned public speaking scores, experiences, and concerns with using AI-PPPs as a learning tool. Implications include recommendations for instructors for improving public speaking instruction with AI-PPPs, as well as developers for refining AI-PPP tools.

Background: Social annotation has emerged as a promising educational technology that fosters collaborative reading and discussion of digital resources among learners. While the positive impact of social annotation on students' learning process and performance is widely acknowledged, students' behavioural patterns in social annotation are underexplored. Objectives: This study investigated patterns in students' use of annotation and response behaviours in social annotation activities. We also explored how students' performance in the behavioural, cognitive, emotional, and social dimensions varied based on their behavioural patterns. Methods: We recruited 93 undergraduates who were enrolled in an elective course at a large North American University. Students were tasked with collaboratively annotating the class readings uploaded to Perusall, a social annotation platform, over 7 weeks. We used metaclustering to determine the optimal number of clusters pertaining to student behaviours. We compared the differences among clusters across multiple performance dimensions. Results and Conclusions: Two distinct clusters were identified and defined as initiators and responders. We found that responders had significantly longer active reading time and exhibited greater social annotation effort compared to initiators. However, initiators received more peer acknowledgement, as evidenced by higher upvotes. No significant difference was found in cognitive insight between initiators and responders, but responders demonstrated significantly higher cognitive discrepancy. Additionally, there were no significant differences in positive and negative tones between initiators and responders; however, responders displayed higher levels of prosocial behaviours than initiators. This study has significant practical implications regarding promoting students' collaborative learning experience in social annotation. Lay Description: What is already known about this topic: Annotation and response behaviours are two primary actions in social annotation.Understanding how students navigate through annotations and respond to their peers' contributions is essential for optimizing their learning experience.Social annotation outcomes can be assessed in the behavioural, cognitive, emotional, and social dimensions. What this paper adds: This paper revealed students' behavioural patterns in social annotation activities.This study offered a comprehensive understanding of the various dimensions of performance among students with different behavioural tendencies. Implications for practice and/or policy: The two student clusters, initiators and responders, revealed distinct engagement patterns in social annotation and informed the design of targeted scaffoldings.Responders were not passive learners since they demonstrated significantly longer active reading time, greater social annotation effort, and higher level of prosocial behaviours.Educators might not need to place significant emphasis on monitoring participants' emotional expressions in social annotation.Social annotation platforms should incorporate features that encourage and reward both initiation and response behaviours. [ABSTRACT FROM AUTHOR]

Purpose: Automatic modulation recognition (AMR) is a challenging problem in intelligent communication systems and has wide application prospects. At present, although many AMR methods based on deep learning have been proposed, the methods proposed by these works cannot be directly applied to the actual wireless communication scenario, because there are usually two kinds of dilemmas when recognizing the real modulated signal, namely, long sequence and noise. This paper aims to effectively process in-phase quadrature (IQ) sequences of very long signals interfered by noise. Design/methodology/approach: This paper proposes a general model for a modulation classifier based on a two-layer nested structure of long short-term memory (LSTM) networks, called a two-layer nested structure (TLN)-LSTM, which exploits the time sensitivity of LSTM and the ability of the nested network structure to extract more features, and can achieve effective processing of ultra-long signal IQ sequences collected from real wireless communication scenarios that are interfered by noise. Findings: Experimental results show that our proposed model has higher recognition accuracy for five types of modulation signals, including amplitude modulation, frequency modulation, gaussian minimum shift keying, quadrature phase shift keying and differential quadrature phase shift keying, collected from real wireless communication scenarios. The overall classification accuracy of the proposed model for these signals can reach 73.11%, compared with 40.84% for the baseline model. Moreover, this model can also achieve high classification performance for analog signals with the same modulation method in the public data set HKDD_AMC36. Originality/value: At present, although many AMR methods based on deep learning have been proposed, these works are based on the model's classification results of various modulated signals in the AMR public data set to evaluate the signal recognition performance of the proposed method rather than collecting real modulated signals for identification in actual wireless communication scenarios. The methods proposed in these works cannot be directly applied to actual wireless communication scenarios. Therefore, this paper proposes a new AMR method, dedicated to the effective processing of the collected ultra-long signal IQ sequences that are interfered by noise. [ABSTRACT FROM AUTHOR]

Sequences with optimal autocorrelation properties play an important role in wireless communication, radar and cryptography. Interleaving is a very important method in constructing the optimal autocorrelation sequence. Tang and Gong gave three different constructions of interleaved sequences (generalized GMW sequences, twin prime sequences and Legendre sequences). Su et al. constructed a series of sequences with optimal autocorrelation magnitude via interleaving Ding-Helleseth-Lam sequences. In this paper we further study the correlation properties of interleaved Legendre sequences and Ding-Helleseth-Lam sequences. [ABSTRACT FROM AUTHOR]

In call centers, the service level is an important metric to measure the reasonability of the staffing schedule. Traditional service level calculation methods are based on the queue theory, which has very strict restrictions and is not suitable for real scenarios. Therefore, in this paper, a data‐driven method to solve the service level prediction problem is proposed to be used. To this end, the relationship between service level and other factors, such as number of calls, number of agents, time, is explored. Then some features are extracted based on empirical analyses and propose to use decision tree based ensemble methods, like random forest and GBDT, to model the relationship between service level and input features. Finally, extensive experimental results show that the proposed method outperforms other baselines significantly. Especially compared with the traditional queue theory methods, our method improves the performance by 6% and 9% in terms of MAE and MAPE. [ABSTRACT FROM AUTHOR]

Nowadays, many location-based applications require the ability of querying $k$ -nearest neighbors over a very large scale of moving objects in road networks, e.g., taxi-calling and ride-sharing services. Traditional grid index with equal-sized cells can not adapt to the skewed distribution of moving objects in real scenarios. Thus, to obtain the fast querying response time, the grid needs to be split into more smaller cells which introduces the side-effect of higher memory cost, i.e., maintaining such a large volume of cells requires a much larger memory space at the server side. In this paper, we present $\sf {SIMkNN}$ , a scalable and in-memory k NN query processing technique. $\sf {SIMkNN}$ is dual-index driven, where we adopt a R-tree to store the topology of the road network and a hierarchical grid model to manage the moving objects in non-uniform distribution. To answer a kNN query in real time, $\sf {SIMkNN}$ adopts the strategy that incrementally enlarges the search area for network distance based nearest neighbor evaluation. It is far from trivial to perform the space expansion within the hierarchical grid index. For a given cell, we first define its neighbors in different directions, then propose a cell communication technique which allows each cell in the hierarchical grid index to be aware of its neighbors at any time. Accordingly, an efficient space expansion algorithm to generate the estimation area is proposed. The experimental evaluation shows that $\sf {SIMkNN}$ outperforms the baseline algorithm in terms of time and memory efficiency. [ABSTRACT FROM AUTHOR]