Abstract-Existing research efforts into tennis visualization have primarily focused on using ball and player tracking data to enhance professional tennis broadcasts and to aid coaches in helping their students. Gathering and analyzing this data typically requires the use of an array of synchronized cameras, which are expensive for non-professional tennis matches. In this paper, we propose TenniVis, a novel tennis match visualization system that relies entirely on data that can be easily collected, such as score, point outcomes, point lengths, service information, and match videos that can be captured by one consumer-level camera. It provides two new visualizations to allow tennis coaches and players to quickly gain insights into match performance. It also provides rich interactions to support ad hoc hypothesis development and testing. We first demonstrate the usefulness of the system by analyzing the 2007 Australian Open men's singles final. We then validate its usability by two pilot user studies where two college tennis coaches analyzed the matches of their own players. The results indicate that useful insights can quickly be discovered and ad hoc hypotheses based on these insights can conveniently be tested through linked match videos.

Abstract-Previous studies on E-transaction time-series have mainly focused on finding temporal trends of transaction behavior. Interesting transactions that are time-stamped and situation-relevant may easily be obscured in a large amount of information. This paper proposes a visual analytics system, Visual Analysis of E-transaction Time-Series (VAET), that allows the analysts to interactively explore large transaction datasets for insights about time-varying transactions. With a set of analyst-determined training samples, VAET automatically estimates the saliency of each transaction in a large time-series using a probabilistic decision tree learner. It provides an effective time-of-saliency (TOS) map where the analysts can explore a large number of transactions at different time granularities. Interesting transactions are further encoded with KnotLines, a compact visual representation that captures both the temporal variations and the contextual connection of transactions. The analysts can thus explore, select, and investigate knotlines of interest. A case study and user study with a real E-transactions dataset (26 million records) demonstrate the effectiveness of VAET.

No abstract

Serendipitous drug usage refers to the unexpected relief of comorbid diseases or symptoms when taking a medication for a different known indication. Historically, serendipity has contributed significantly to identifying many new drug indications. If patient-reported serendipitous drug usage in social media could be computationally identified, it could help generate and validate drug-repositioning hypotheses. We investigated deep neural network models for mining serendipitous drug usage from social media. We used the word2vec algorithm to construct wordembedding features from drug reviews posted in a WebMD patient forum. We adapted and redesigned the convolutional neural network, long short-term memory network, and convolutional long short-term memory network by adding contextual information extracted from drug-review posts, information-filtering tools, medical ontology, and medical knowledge. We trained, tuned, and evaluated our models with a gold-standard dataset of 15,714 sentences (447 [2.8%] describing serendipitous drug usage). Additionally, we compared our deep neural networks to support vector machine, random forest, and AdaBoost.M1 algorithms. Context information helped reduce the false-positive rate of deep neural network models. If we used an extremely imbalanced dataset with limited instances of serendipitous drug usage, deep neural network models did not outperform other machine-learning models with n-gram and context features. However, deep neural network models could more effectively use word embedding in feature construction, an advantage that makes them worthy of further investigation. Finally, we implemented naturallanguage processing and machine-learning methods in a web-based application to help scientists and software developers mine social media for serendipitous drug usage.

No abstract