We describe our system (DT Team) submitted at SemEval-2017 Task 1, Semantic Textual Similarity (STS) challenge for English (Track 5). We developed three different models with various features including similarity scores calculated using word and chunk alignments, word/sentence embeddings, and Gaussian Mixture Model (GMM). The correlation between our system's output and the human judgments were up to 0.8536, which is more than 10% above baseline, and almost as good as the best performing system which was at 0.8547 correlation (the difference is just about 0.1%). Also, our system produced leading results when evaluated with a separate STS benchmark dataset. The word alignment and sentence embeddings based features were found to be very effective.
We present in this paper an automated method to assess the quality of Jupyter notebooks. The quality of notebooks is assessed in terms of reproducibility and executability. Specifically, we automatically extract a number of expert-defined features for each notebook, perform a feature selection step, and then trained supervised binary classifiers to predict whether a notebook is reproducible and executable, respectively. We also experimented with semantic code embeddings to capture the notebooks' semantics. We have evaluated these methods on a dataset of 306,539 notebooks and achieved an F1 score of 0.87 for reproducibility and 0.96 for executability (using expert-defined features) and an F1 score of 0.81 for reproducibility and 0.78 for executability (using code embeddings). Our results suggest that semantic code embeddings can be used to determine with good performance the reproducibility and executability of Jupyter notebooks, and since they can be automatically derived, they have the advantage of no need for expert involvement to define features.
We present a novel approach to intro-to-programming domain model discovery from textbooks using an over-generation and ranking strategy. We first extract candidate key phrases from each chapter in a Computer Science textbook focusing on intro-to-programming and then rank those concepts according to a number of metrics such as the standard tf-idf weight used in information retrieval and metrics produced by other text ranking algorithms. Specifically, we conduct our work in the context of developing an intelligent tutoring system for source code comprehension for which a specification of the key programming concepts is needed - the system monitors students' performance on those concepts and scaffolds their learning process until they show mastery of the concepts. Our experiments with programming concept instruction from Java textbooks indicate that the statistical methods such as KP Miner method are quite competitive compared to other more sophisticated methods. Automated discovery of domain models will lead to more scalable Intelligent Tutoring Systems (ITSs) across topics and domains, which is a major challenge that needs to be addressed if ITSs are to be widely used by millions of learners across many domains.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.