Azza Abouzied

One LED is Enough: Catalyzing Face-to-face Interactions at Conferences with a Gentle Nudge

Face-to-face social interactions among strangers today are becoming increasingly rare as people turn towards computer-mediated networking tools. Todays tools, however, are based on the following assumptions: increased information encourages interaction, profiles are good representations of users to other users, and computer-mediated communications prior to face-to-face meetings lead to better outcomes. This paper describes CommonTies, a gentle technological in the form of a wearable accessory, that encourages immediate, face-to-face, organic social interactions among strangers at conferences. By not exposing any profile information, CommonTies preserves an element of mystery and enables self-disclosure of information through conversation. We evaluate our system through a field study at a three-day research conference - CSCW 2014. We find that despite our information-scarce design, users were willing to interact with strangers and 74% of the interactions initiated by CommonTies were reported as novel and useful.

Scalable package queries in relational database systems

Traditional database queries follow a simple model: they define constraints that each tuple in the result must satisfy. This model is computationally efficient, as the database system can evaluate the query conditions on each tuple individually. However, many practical, real-world problems require a collection of result tuples to satisfy constraints collectively, rather than individually. In this paper, we present package queries, a new query model that extends traditional database queries to handle complex constraints and preferences over answer sets. We develop a full-fledged package query system, implemented on top of a traditional database engine. Our work makes several contributions. First, we design PaQL, a SQL-based query language that supports the declarative specification of package queries. We prove that PaQL is at least as expressive as integer linear programming, and therefore, evaluation of package queries is in general NP-hard. Second, we present a fundamental evaluation strategy that combines the capabilities of databases and constraint optimization solvers to derive solutions to package queries. The core of our approach is a set of translation rules that transform a package query to an integer linear program. Third, we introduce an offline data partitioning strategy allowing query evaluation to scale to large data sizes. Fourth, we introduce SketchRefine, a scalable algorithm for package evaluation, with strong approximation guarantees ((1 ± e)6-factor approximation). Finally, we present extensive experiments over real-world and benchmark data. The results demonstrate that SketchRefine is effective at deriving high-quality package results, and achieves runtime performance that is an order of magnitude faster than directly using ILP solvers over large datasets.

Codo: Fundraising with Conditional Donations

Crowdfunding websites like Kickstarter and Indiegogo offer project organizers the ability to market, fund, and build a community around their campaign. While offering support and flexibility for organizers, crowdfunding sites provide very little control to donors. In this paper, we investigate the idea of empowering donors by allowing them to specify conditions for their crowdfunding contributions. We introduce a crowdfunding system, Codo, that allows donors to specify conditional donations. Codo allow donors to contribute to a campaign but hold off on their contribution until certain specific conditions are met (e.g. specific members or groups contribute a certain amount). We begin with a micro study to assess several specific conditional donations based on their comprehensibility and usage likelihood. Based on this study, we formalize conditional donations into a general grammar that captures a broad set of useful conditions. We demonstrate the feasibility of resolving conditions in our grammar by elegantly transforming conditional donations into a system of linear inequalities that are efficiently resolved using off-the-shelf linear program solvers. Finally, we designed a user-friendly crowdfunding interface that supports conditional donations for an actual fund raising campaign and assess the potential of conditional donations through this campaign. We find preliminary evidence that roughly 1 in 3 donors make conditional donations and that conditional donors donate more compared to direct donors.

PackageBuilder: from tuples to packages

In this demo, we present PackageBuilder, a system that extends database systems to support package queries. A package is a collection of tuples that individually satisfy base constraints and collectively satisfy global constraints. The need for package support arises in a variety of scenarios: For example, in the creation of meal plans, users are not only interested in the nutritional content of individual meals (base constraints), but also care to specify daily consumption limits and control the balance of the entire plan (global constraints). We introduce PaQL, a declarative SQL-based package query language, and the interface abstractions which allow users to interactively specify package queries and easily navigate through their results. To efficiently evaluate queries, the system employs pruning and heuristics, as well as state-of-the-art constraint optimization solvers. We demonstrate PackageBuilder by allowing attendees to interact with the systems interface, to define PaQL queries and to observe how query evaluation is performed.

Synthesizing Extraction Rules from User Examples with SEER

Our demonstration showcases SEERs end-to-end Information Extraction (IE) workflow where users highlight texts they wish to extract. Given a small set of user-specified example extractions, SEER synthesizes easy-to-understand IE rules and suggests them to the user. In addition to rule suggestions, users can quickly pick the desired rule by filtering the rule suggestion by accepting or rejecting proposed extractions. SEERs workflow allows users to jump start the IE rule development cycle; it is a less time-consuming alternative to machine learning methods that require large labeled datasets or rule-based approaches that are labor-intensive. SEERs design principles and learning algorithm are motivated by how rule developers naturally construct data extraction rules.

Shelter Dynamics in Refugee and IDP Camps: Customization, Permanency, and Opportunities

The UNHCR estimates that the average forced displacement period is 17 years, which many refugees and IDPs (Internally Displaced Persons) spend entirely in camps. This reality has caused camps to be increasingly considered as permanent cities of our future rather than temporary relief solutions. Unfortunately, this recognition has not been matched by corresponding increases in the planning or resources devoted to camps. In the case of shelter, a basic human need, little to no architectural infrastructure exists and urban planning remains short-term. As a result, camp dwellers are often forced to take it upon themselves to transform existing humanitarian storage facilities into essential domiciles, markets, and communities. In this paper, we describe our observations and survey results on the state of and practices surrounding shelter from three camps in north Iraq. Our findings illustrate the various modes of shelter that exist due to economic and political expediency, and highlight opportunities for ICTs to improve the quality of life for millions of displaced residents.

Harnessing data loss with forgetful data structures

Forgetting, losing, or corrupting data is almost universally considered harmful in computer science and blasphemous in database and file systems. Typically, loss of data is a consequence of unmanageable or unexpected lower layer deficiencies that the user process must be protected from through multiple layers of storage abstractions and redundancies. We argue that forgetfulness can be a resource for system design and that, like durability, security or integrity, can be used to achieve uncommon, but potentially important goals such as privacy, plausible deniability, and the right to be forgotten. We define the key properties of forgetfulness and draw inspiration from human memory. We develop a data structure, the forgit, that can be used to store images, audio files, videos or numerical data and eventually forget. Forgits are a natural data store for a multitude of todays cloud-based applications and we discuss their use, effectiveness, and limitations in this paper.

Expressive Time Series Querying with Hand-Drawn Scale-Free Sketches

We present Qetch, a tool where users freely sketch patterns on a scale-less canvas to query time series data without specifying query length or amplitude. We study how humans sketch time series patterns --- humans preserve visually salient perceptual features but often non-uniformly scale and locally distort a pattern --- and we develop a novel matching algorithm that accounts for human sketching errors. Qetch enables the easy construction of complex and expressive queries with two key features: regular expressions over sketches and relative positioning of sketches to query multiple time-aligned series. Through user studies, we demonstrate the effectiveness of Qetchs different interaction features. We also demonstrate the effectiveness of Qetchs matching algorithm compared to popular algorithms on targeted, and exploratory query-by-sketch search tasks on a variety of data sets.

SEER: Auto-Generating Information Extraction Rules from User-Specified Examples

Time-consuming and complicated best describe the current state of the Information Extraction (IE) field. Machine learning approaches to IE require large collections of labeled datasets that are difficult to create and use obscure mathematical models, occasionally returning unwanted results that are unexplainable. Rule-based approaches, while resulting in easy-to-understand IE rules, are still time-consuming and labor-intensive. SEER combines the best of these two approaches: a learning model for IE rules based on a small number of user-specified examples. In this paper, we explain the design behind SEER and present a user study comparing our system against a commercially available tool in which users create IE rules manually. Our results show that SEER helps users complete text extraction tasks more quickly, as well as more accurately.

From Alley to Landfill: Challenges of and Design Opportunities for Cleaning Dhaka's Communal Trash

Garbage is an endemic problem in developing cities due to the continual influx of migrants from rural areas coupled with deficient municipal capacity planning. In cities like Dhaka, open waste dumps contribute to the prevalence of disease, environmental contamination, catastrophic flooding, and deadly fires. Recent interest in the garbage problem has prompted cursory proposals to introduce technology solutions for mapping and fundraising. Yet, the role of technology and its potential benefits are unexplored in this large-scale problem. In this paper, we contribute to the understanding of the waste ecology in Dhaka and how the various actors acquire, perform, negotiate, and coordinate their roles. Within this context, we explore design opportunities for using computing technologies to support collaboration between waste pickers and residents of these communities. We find opportunities in the presence of technology and the absence of mechanisms to facilitate coordination of community funding and crowd work.