Interactive tools help users filter relevant information from massive online sources, like news feeds and online discussion forums, by enabling them to externalize their preferences. However, users’ information goals and preferences are often complex and are comprised of data attributes and a user’s subjective judgements over these attributes. For instance, when filtering news articles based on their newsworthiness, the system must capture both data attributes like recency and shareability of the article, along with the user’s personal and flexible assessment of news sentiment. While most interactive tools enable users to externalize goals that are expressible as true/false statements, they do not support incorporating subjective, loosely structured judgements of data attributes which fulfill complex goals. In this paper, we introduce Teachable Facets (TF), widgets that users can create on the fly to filter relevant information to improve the sense-making of analysts. These teachable widgets employ a Machine Teaching (MT) framework to enable users to formulate personalized filtering criteria for complex, multi-dimensional, loosely indexed, and unstructured data; teach a filtering criterion using representative samples; apply these filters to new data streams; and assess the relevance of outcomes. Through a user study, we evaluate the performance of these filters based on their ability to discover relevant items and the expressibility they offer to the users in teaching criteria. In our discussion, we identify ways this approach might improve future systems and delineate implications should such systems be deployed broadly.

Interactive interfaces in tandem with Machine Learning (ML) models support user understanding of model uncertainty, build confidence, improve predictive accuracy and enable users to teach application-specific concepts that are difficult for the model to learn otherwise. These systems offer empirically proven benefits due to tightly coupled feedback loops and workflow scaffolding. However, deployment with ML non-experts who cannot manage the complex, expertise-heavy process remains challenging. Through deployment with non-expert users in a common classification task, we investigate the impact of human factors of machine teaching interfaces such as user expectations, their perceptions of the learning process and user engagement with respect to teaching process and outcomes. We measure how affective and performance attributes shape the success or failure of the process. Finally, we reflect on how intelligent user interfaces can be designed to accommodate these factors for successful deployment with a broad spectrum of human adjudicators.

Interactive machine learning (iML) tools help to make ML accessible to users with limited ML expertise. However, gathering necessary training data and expertise for model-building remains challenging. Transfer learning, a process where learned representations from a model trained on potentially terabytes of data can be transferred to a new, related task, offers the possibility of providing ”building blocks” for non-expert users to quickly and effectively apply ML in their work. However, transfer learning largely remains an expert tool due to its high complexity. In this paper, we design a prototype to understand non-expert user behavior in an interactive environment that supports transfer learning. Our findings reveal a series of data- and perception-driven decision-making strategies non-expert users employ, to (in)effectively transfer elements using their domain expertise. Finally, we synthesize design implications which might inform future interactive transfer learning environments.

An important challenge in building explainable artificially intelligent (AI) systems is designing interpretable explanations. AI models often use low-level data features which may be hard for humans to interpret. Recent research suggests that situating machine decisions in abstract, human understandable concepts can help. However, it is challenging to determine the right level of conceptual mapping. In this research, we explore granularity (of data features) and context (of data instances) as dimensions underpinning conceptual mappings. Based on these measures, we explore strategies for designing explanations in classification models. We introduce an end-to-end concept elicitation pipeline that supports gathering high-level concepts for a given data set. Through crowd-sourced experiments, we examine how providing conceptual information shapes the effectiveness of explanations, finding that a balance between coarse and fine-grained explanations help users better estimate model predictions. We organize our findings into systematic themes that can inform design considerations for future systems.

Visualization tools facilitate exploratory data analysis, but fall short at supporting hypothesis-based reasoning. We conducted an exploratory study to investigate how visualizations might support a concept-driven analysis style, where users can optionally share their hypotheses and conceptual models in natural language, and receive customized plots depicting the fit of their models to the data. We report on how participants leveraged these unique affordances for visual analysis. We found that a majority of participants articulated meaningful models and predictions, utilizing them as entry points to sensemaking. We contribute an abstract typology representing the types of models participants held and externalized as data expectations. Our findings suggest ways for rearchitecting visual analytics tools to better support hypothesis- and model-based reasoning, in addition to their traditional role in exploratory analysis. We discuss the design implications and reflect on the potential benefits and challenges involved.

Engaging museum visitors in data exploration using full-body interaction is still a challenge. In this paper, we explore four strategies for providing entry-points to the interaction: instrumenting the floor; forcing collaboration; implementing multiple body movements to control the same effect; and, visualizing the visitors' silhouette beside the data visualization. We discuss preliminary results of an in-situ study with 56 museum visitors at Discovery Place, and provide design recommendations for crafting engaging Human-Data Interaction experiences.

Motivation: Rapid advancement in high throughput genome and transcriptome sequencing (HTS) and mass spectrometry (MS) technologies has enabled the acquisition of the genomic, transcriptomic and proteomic data from the same tissue sample. We introduce a computational framework, ProTIE, to integratively analyze all three types of omics data for a complete molecular profile of a tissue sample. Our framework features MiStrVar, a novel algorithmic method to identify micro structural variants (microSVs) on genomic HTS data. Coupled with deFuse, a popular gene fusion detection method we developed earlier, MiStrVar can accurately profile structurally aberrant transcripts in tumors. Given the breakpoints obtained by MiStrVar and deFuse, our framework can then identify all relevant peptides that span the breakpoint junctions and match them with unique proteomic signatures. Observing structural aberrations in all three types of omics data validates their presence in the tumor samples. Results: We have applied our framework to all The Cancer Genome Atlas (TCGA) breast cancer Whole Genome Sequencing (WGS) and/or RNA-Seq datasets, spanning all four major subtypes, for which proteomics data from Clinical Proteomic Tumor Analysis Consortium (CPTAC) have been released. A recent study on this dataset focusing on SNVs has reported many that lead to novel peptides. Complementing and significantly broadening this study, we detected 244 novel peptides from 432 candidate genomic or transcriptomic sequence aberrations. Many of the fusions and microSVs we discovered have not been reported in the literature. Interestingly, the vast majority of these translated aberrations, fusions in particular, were private, demonstrating the extensive inter-genomic heterogeneity present in breast cancer. Many of these aberrations also have matching out-of-frame downstream peptides, potentially indicating novel protein sequence and structure. Availability and implementation: MiStrVar is available for download at https://bitbucket.org/compbio/mistrvar, and ProTIE is available at https://bitbucket.org/compbio/protie. Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.