Educational question generation (EQG) is a crucial component of intelligent educational systems, significantly aiding self-assessment, active learning, and personalized education. While EQG systems have emerged, existing datasets typically rely on predefined, carefully edited texts, failing to represent real-world classroom content, including lecture speech with a set of complementary slides. To bridge this gap, we collect a dataset of educational questions based on videos from real-world lectures. On this realistic dataset, we find that current methods for EQG struggle to accurately generate questions from educational videos, particularly in aligning with specific timestamps and target answers. Common challenges include selecting informative contexts from extensive transcripts and ensuring generated questions meaningfully incorporate the target answer. To address the challenges, we introduce a novel framework utilizing large language models (LLMs) for dynamically selecting and rewriting contexts based on target timestamps and answers in lecture videos. First, our framework selects contexts from both lecture transcripts and video keyframes based on answer relevance and temporal proximity. Then, we integrate the contexts selected from both modalities and rewrite them into answer-containing knowledge statements, to enhance the logical connection between the contexts and the desired answer. Quantitative evaluation and human evaluation show that our approach improves the quality and relevance of the generated questions.

While the Question Generation (QG) task has been increasingly adopted in educational assessments, its evaluation remains limited by approaches that lack a clear connection to the educational values of test items.In this work, we introduce test item analysis, a method frequently used by educators to assess test question quality, into QG evaluation.Specifically, we construct pairs of candidate questions that differ in quality across dimensions such as topic coverage, item difficulty, item discrimination, and distractor efficiency.We then examine whether existing QG evaluation approaches can effectively distinguish these differences.Our findings reveal significant shortcomings in these approaches with respect to accurately assessing test item quality in relation to student performance.To address this gap, we propose a novel QG evaluation framework, QG-SMS, which leverages Large Language Model for Student Modeling and Simulation to perform test item analysis.As demonstrated in our extensive experiments and human evaluation study, the additional perspectives introduced by the simulated student profiles lead to a more effective and robust assessment of test items.

This paper introduces a large-scale human-labeled dataset for the Vietnamese POS tagging task on conversational texts. To this end, wepropose a new tagging scheme (with 36 POS tags) consisting of exclusive tags for special phenomena of conversational words, developthe annotation guideline and manually annotate 16.310K sentences using this guideline. Based on this corpus, a series of state-of-the-art tagging methods has been conducted to estimate their performances. Experimental results showed that the Conditional Random Fields model using both automatically learnt features from deep neural networks and handcrafted features yielded the best performance. Thismodel achieved 93.36% in the accuracy score which is 1.6% and 2.7% higher than the model using either handcrafted features orautomatically-learnt features, respectively. This result is also a little bit higher than the model of fine-tuning BERT by 0.94% in theaccuracy score. The performance measured on each POS tag is also very high with >90% in the F1 score for 20 POS tags and >80%in the F1 score for 11 POS tags. This work provides the public dataset and preliminary results for follow-up research on this interesting direction.

This paper investigates the ways in which entrepreneurs use communities of practice (CoPs) to express themselves, using narrative theory and rhetorical analysis, to gain insight into an electronic social network medium, namely, YoungEntrepreneur.com. In particular, the study focuses on CoPs themes, including why entrepreneurs engage in CoPs, what role the moderators and resident entrepreneurs can play in managing online CoPs, on communication rituals of the knowledge sharing through interactivity, and on ‘how to develop an intervention’ to maintain and stimulate entrepreneurs for engaging in an online community. Findings reveal that the topic title plays a major role in attracting people. Successful topics with successful conclusions (in terms of the original query that was answered) will not necessarily get high responses and vice versa. It is observed that the domain expert does not play a big role in keeping the discussion going. Finally, the study also discovered that entrepreneurs like to communicate in a story telling genre. A comprehensive set of engagement measurement tools are introduced to effectively measure the engagement in a virtual CoP, along with a classification to define and categorise discourse of messages in terms of content and context, which allow practitioners to understand the effectiveness of a social networking site.

QoS multicast routing with more than one metric has always been technically challenging, since many of them are NP-hard. Most existing QoS multicast routing algorithms are heuristic. Furthermore, many of them considered only the unicast shortest paths, either based on propagation delay or the number of hops. In this paper, we observed that the end-to-end delay experienced by network packets is determined not only by the propagation distance, but also (and in many cases largely) by the hop-by-hop transmission and queueing delay, which in turn depends on the available bandwidth. Based on this observation we propose an optimal multicast routing algorithm, which maximizes the reserved bandwidth while satisfying both delay and bandwidth constraints. Its correctness and time-complexity of O(n/sup 2/log n) are formally verified, where n is the number of nodes in the network. We also extend the algorithm to be dynamic, such that nodes are allowed to join or leave the network, with the multicast route updated dynamically. We carefully evaluate both the static and the dynamic algorithms via simulation. We found that the new algorithm significantly reduces the end-to-end delay experienced by network packets; its dynamic version is able to rapidly update the multicast route using little computation time. We believe that both versions of the new algorithm are important to the area of QoS-based multicast routing; their design and analysis may give new insights to network researchers in their future design of routing protocols.

Fast transmission among many sources and destinations usually has the goal of minimizing some objective (such as cost) while satisfying some constraints such as delay or bandwidth. Due to these constraints, the solution sometimes is achieved only through the use of multiple-shared multicast trees (MSMT), which use more than one multicast tree to serve multiple sources and destinations. Minimizing the number of multiple-shared multicast trees, however, has been proven to be NP-complete. In this paper we propose three algorithms for this problem and compare them with an existing one. We can prove, theoretically and through simulations, that our algorithms gives the same good results with a significantly shorter run-time. On the other hand, fast transmission and high reliability are also achieved by selecting the path with large bandwidth. We apply the MBDC (maximum bottleneck bandwidth with delay constraint) algorithm, which we previously proposed, to the MSMT algorithms to optimize them in terms of bandwidth. Proof of correctness, estimation of run-time, and simulation results of these extensions are also provided.