My personal highlights among the full paper presentations:
- Fernando Diaz et al. – Search Result Prefetching Using Cursor Movement. To reduce latency during page fetch, the authors investiagte a wide range of cursor movement features that forecast SERP clicks. As the click likelihood exceeds a threshold value, the page is fetched into cache and the user experiences no latency as the actual click occurs.
- Ryan Burton and Kevyn Collins-Thompson – User Behavior in Asynchronous Slow Search. The authors present an asynchronous (slow) search framework in which users are encouraged to identify topics on which they are willing to wait longer in exchange for better result quality. This interaction model allows search engines to use a greater portfolio of retrieval methods than can be achieved in the typical 30-mS round-trip.
- Tetsuya Sakai – Statistical Significance, Power, and Sample Sizes: A Systematic Review of SIGIR and TOIS, 2006-2015. The author present a review of 10 years of IR experimentation in SIGIR and TOIS and systematically analyzes significance tests, sample and effect sizes.
- Chenliang Li et al. – Topic Modeling for Short Texts with Auxiliary Word Embeddings. The authors propose a Dirichlet Multinomial Mixture (DMM) model powered by a Generalized Polya Urn (GPU) sampling process in order to address sparsity and efficiency issues in short text topic modeling.
- Rui Yan et al. – Learning to Respond with Deep Neural Networks for Retrieval-Based Human-Computer Conversation System. The authors describe a deep neural network architecture incorporating feed-forward word embeddings, recurrent LSTMs as well as convolutional layers for the task of conversational agent training.
As well as some promising short papers:
- Tetsuya Sakai –
Two Sample T-tests for IR Evaluation: Student or Welch? The author contests the widely-accepted notion that a Welch t-test be unconditionally preferable over a two-sample Student test. The investigation concludes that, if sample sizes differ substantially, and if the larger sample has a substantially larger variance, Welch’s t-test may not be reliable.
- Bevan Koopman et al. –
A Test Collection for Matching Patients to Clinical Trials. The authors annotated 60 existing TREC CDS patient descriptions in terms of their eligibility for participation in a wide range of publicly advertised clinical trials.
- Sumit Sidana et al. –
Health Monitoring on Social Media over Time. Using a spatio-temporal topic modeling approach, the authors investigate which medical conditions people manifest on social media at different geographical locations as well as points in time (e.g., throughout the seasons).