A5102726483- Mobile Crowdsensing and Crowdsourcing
- Advanced Text Analysis Techniques
- Personal Information Management and User Behavior
- Semantic Web and Ontologies
- Natural Language Processing Techniques
- Information Retrieval and Search Behavior
- Scientific Computing and Data Management
- Software Engineering Research
- Topic Modeling
- Data Visualization and Analytics
- Data Stream Mining Techniques
- Web Data Mining and Analysis
- Speech and dialogue systems
- Biomedical Text Mining and Ontologies
- AI in Service Interactions
- Recommender Systems and Techniques
- Artificial Intelligence in Law
- Wikis in Education and Collaboration
- Data Quality and Management
- Usability and User Interface Design
- Social Media and Politics
- Context-Aware Activity Recognition Systems
- Multi-Agent Systems and Negotiation
- Human Mobility and Location-Based Analysis
- Evaluation and Performance Assessment
Allen Institute
2022-2024
Allen Institute for Artificial Intelligence
2023-2024
Carnegie Mellon University
2016-2021
Human Media
2016-2018
National Hsinchu University of Education
2012
Crowdsourcing provides a scalable and efficient way to construct labeled datasets for training machine learning systems. However, creating comprehensive label guidelines crowdworkers is often prohibitive even seemingly simple concepts. Incomplete or ambiguous can then result in differing interpretations of concepts inconsistent labels. Existing approaches improving quality, such as worker screening detection poor work, are ineffective this problem lead rejection honest work missed...
In our era of rapid technological advancement, the research landscape for writing assistants has become increasingly fragmented across various communities. We seek to address this challenge by proposing a design space as structured way examine and explore multidimensional intelligent interactive assistants. Through large community collaboration, we five aspects assistants: task, user, technology, interaction, ecosystem. Within each aspect, define dimensions (i.e., fundamental components an...
Consumers conducting comparison shopping, researchers making sense of competitive space, and developers looking for code snippets online all face the challenge capturing information they find later use without interrupting their current flow. In addition, during many learning exploration tasks, people need to externalize mental context, such as estimating how urgent a topic is follow up on, or rating piece evidence "pro" "con," which helps scaffold subsequent deeper exploration. However,...
Scientific discoveries are often driven by finding analogies in distant domains, but the growing number of papers makes it difficult to find relevant ideas a single discipline, let alone other domains. To provide computational support for across we introduce SOLVENT, mixed-initiative system where humans annotate aspects research that denote their background (the high-level problems being addressed), purpose specific mechanism (how they achieved purpose), and findings (what learned/achieved),...
Crowd-powered conversational assistants have been shown to be more robust than automated systems, but do so at the cost of higher response latency and monetary costs. A promising direction is combine two approaches for high quality, low latency, solutions. In this paper, we introduce Evorus, a crowd-powered assistant built automate itself over time by (i) allowing new chatbots easily integrated scenarios, (ii) reusing prior crowd answers, (iii) learning automatically approve candidates. Our...
When reading a scholarly article, inline citations help researchers contextualize the current article and discover relevant prior work. However, it can be challenging to prioritize make sense of hundreds encountered during literature reviews. This paper introduces CiteSee, tool that leverages user's publishing, reading, saving activities provide personalized visual augmentations context around citations. First, CiteSee connects familiar contexts by surfacing known user had cited or opened....
Scholars who want to research a scientific topic must take time read, extract meaning, and identify connections across many papers. As literature grows, this becomes increasingly challenging. Meanwhile, authors summarize prior in papers' related work sections, though is scoped support single paper. A formative study found that while reading multiple paragraphs helps overview topic, it hard navigate overlapping diverging references foci. In work, we design system, Relatedly, scaffolds...
Crowdsourcing offers a powerful new paradigm for online work. However, real world tasks are often interdependent, requiring big picture view of the difference pieces involved. Existing crowdsourcing approaches that support such -- ranging from Wikipedia to flash teams bottlenecked by relying on small number individuals maintain picture. In this paper, we explore idea computational system can scaffold an emerging entirely through contributions individuals, each whom sees only part whole. To...
Whether figuring out where to eat in an unfamiliar city or deciding which apartment live in, consumer generated data (i.e. reviews and forum posts) are often important influence online decision making. To make sense of these rich repositories diverse opinions, searchers need sift through a large number characterize each item based on aspects that they care about. We introduce novel system, SearchLens, build up collection "Lenses" reflect their different latent interests, compose the Lenses...
Reviewing the literature to understand relevant threads of past work is a critical part research and vehicle for learning. However, as scientific grows challenges users find make sense many different grow well. Previous has helped scholars group papers with citation information or textual similarity using standalone tools overview visualizations. Instead, in this we explore tool integrated into users' reading process that helps them leveraging authors' existing summarization threads,...
Crowdsourced clustering approaches present a promising way to harness deep semantic knowledge for complex information. However, existing have difficulties supporting the global context needed workers generate meaningful categories, and are costly because all items require human judgments. We introduce Alloy, hybrid approach that combines richness of judgments with power machine algorithms. Alloy supports greater through new "sample search" crowd pattern which changes crowd's task from...
Efficiently reviewing scholarly literature and synthesizing prior art are crucial for scientific progress. Yet, the growing scale of publications burden knowledge make synthesis research threads more challenging than ever. While significant has been devoted to helping scholars interact with individual papers, building scattered across multiple papers remains a challenge. Most top-down (and LLMs) it difficult personalize iterate on output, while bottom-up is costly in time effort. Here, we...
People spend a significant amount of time trying to make sense the internet, collecting content from variety sources and organizing it decisions achieve their goals. While humans are able fluidly iterate on information in minds, existing tools approaches introduce friction into process. We Fuse, browser extension that externalizes users' working memory by combining low-cost collection with lightweight organization compact card-based sidebar is always available. Fuse helps users...
While there is an enormous amount of information online for making decisions such as choosing a product, restaurant, or school, it can be costly users to synthesize that into confident decisions. Information users' many different criteria needs gathered from sources structure where they compared and contrasted. The usefulness each criterion differentiating potential options opaque users, evidence reviews may subjective conflicting, requiring interpret under their personal context. We...
In order to help scholars understand and follow a research topic, significant has been devoted creating systems that discover relevant papers authors. Recent approaches have shown the usefulness of highlighting authors while engage in paper discovery. However, these do not capture utilize users' evolving knowledge We reflect on design space introduce ComLittee, literature discovery system supports author-centric exploration. contrast paper-centric interaction prior systems, ComLittee's...
Research consumption has been traditionally limited to the reading of academic papers-a static, dense, and formally written format. Alternatively, pre-recorded conference presentation videos, which are more dynamic, concise, colloquial, have recently become widely available but potentially under-utilized. In this work, we explore design space benefits for combining papers talk videos leverage their complementary nature provide a rich fluid research experience. Based on formative co-design...
With the rapid growth of scholarly archives, researchers subscribe to "paper alert'' systems that periodically provide them with recommendations recently published papers are similar previously collected papers. However, sometimes struggle make sense nuanced connections between recommended and their own research context, as existing only present paper titles abstracts. To help spot these connections, we PaperWeaver, an enriched alerts system provides contextualized text descriptions based on...
Mixed language data is one of the difficult yet less explored domains natural processing. Most research in fields like machine translation or sentiment analysis assume monolingual input. However, people who are capable using more than often communicate multiple languages at same time. Sociolinguists believe this "code-switching" phenomenon to be socially motivated. For example, express solidarity establish authority. past work depend on external tools resources, such as part-of-speech...
Patients researching medical diagnoses, scientist exploring new fields of literature, and students learning about domains are all faced with the challenge capturing information they find for later use. However, saving is challenging on mobile devices, where small screen font sizes combined inaccuracy finger based touch screens makes it time consuming stressful people to select save text future Furthermore, beyond simply selecting a region bounded device, in many data exploration tasks...
People engaged in complex searches such as planning a vacation or understanding their medical symptoms are often overwhelmed by opening and managing many tabs. These challenges exacerbated search moves to smartphones mobile devices where screen real-estate is limited tasks frequently suspended, resumed, interleaved. Rather than continue utilize tab-based browsing for search, we introduce new way of through scaffolded interface. The list results serves mutable workspace, user can track...
Scholarly publications are key to the transfer of knowledge from scholars others. However, research papers information-dense, and as volume scientific literature grows, need for new technology support reading process grows. In contrast finding papers, which has been transformed by Internet technology, experience changed little in decades. The PDF format sharing is widely used due its portability, but it significant downsides including: static content, poor accessibility low-vision readers,...
In communities with social hierarchies, fear of judgment can discourage communication. While anonymity may alleviate some pressure, fully anonymous spaces enable toxic behavior and hide the context that motivates people to participate helps them tailor their We explore a design space meronymous communication, where reveal carefully chosen aspects identity also leverage trusted endorsers gain credibility. implemented these ideas in system for scholars meronymously seek receive paper...
Scholarly publications are key to the transfer of knowledge from scholars others. However, research papers information-dense, and as volume scientific literature grows, greater need for new technology support scholars. In contrast process finding papers, which has been transformed by Internet technology, experience reading changed little in decades. For instance, PDF format sharing remains widely used due its portability but significant downsides, inter alia, static content poor...
Tabs have become integral to browsing the Web yet changed little since their introduction nearly 20 years ago. In contrast, internet has gone through dramatic changes, with users increasingly moving from navigating websites exploring information across many sources support online sensemaking. This paper investigates how tabs today are overloaded a diverse set of functionalities and issues face when managing them. We interviewed ten workers asking about tab management strategies walk each...
Despite the increasing complexity and scale of people's online activities, browser interfaces have stayed largely same since tabs were introduced in major browsers nearly 20 years ago. The gap between simple tab-based users' tasks can lead to serious adverse effects – commonly referred as "tab overload." This paper introduces a Chrome extension called Tabs.do, which explores bringing task-centric approach browser, helping users group their into then organize, prioritize, switch those...