Working with real-world data has increasingly become a popular context for introductory computing courses. As a valuable 21st century skill, preparing students to be able to divine meaning from data can be useful to their long-term careers. Because Data Science aligns so closely with computing, many of the topics and problems it affords as a context can support the core learning objectives in introductory computing classes. In many instances, incorporating a real-world dataset to provide concrete context for an activity or assignment can improve student engagement and understanding of the abstract educational content being presented. However, there are many problems inherent to bringing real-world data into introductory courses. How do instructors, with finite amounts of time and energy, find and prepare suitable datasets for their pedagogical needs? Once the datasets are ready, how can students conveniently interact with and draw meaning from the datasets, especially when they are used in complex projects that are typical of later introductory courses? On the other hand, how does an instructor balance the complexities of using real-world datasets in the classroom, making sure that students appreciate the meaningfulness of course activities and their connection to learning objectives? This panel brings together experts with experience in using real-world data in introductory computing courses. Each panelist provides unique perspectives and skills to the problem of preparing, interacting, visualizing, and using pedagogical datasets. This panel should be of particular interest to instructors who are considering integrating current and real-world data into their assignments and projects, and to educational developers who want to create and manage datasets for pedagogical purposes. The panel will follow a conventional format: 5 minutes of introduction, 10 minutes for each panelist to present, and then 30 minutes for audience Q&A.

This paper presents work on a code framework and methodology to facilitate the introduction of large, real-time, online data sources into introductory (or advanced) Computer Science courses. The framework is generic in the sense that no prior scaffolding or template specification is needed to make the data accessible, as long as the source uses a standard format such as XML, CSV, or JSON. The implementation described here maintains minimal syntactic overhead while relieving novice programmers from low-level issues of parsing raw data from a web-based data source. It interfaces directly with data structures and representations defined by the students themselves, rather than predefined and supplied by the library. Together, these features allow students and instructors to focus on algorithmic aspects of processing a wide variety of live and large data sources, without having to deal with low-level connection, parsing, extraction, and data binding. The library, available at http://cs.berry.edu/big-data, has been used in an introductory programming course based on Processing.

A number of contextualized approaches to teaching introductory Computer Science (CS) courses have been developed in the past few years, catering to students with different interests and learning styles. For instance, entire courses have been developed around media computation or robots (real and virtual). There is however one context which, to our knowledge, has not been exploited in a systematic fashion - that of "big data," by which we mean massive, openly accessible online datasets from a wide variety of sources. We present progress on a code framework and methodology to facilitate the incorporation of large, online data sets into traditional CS1 and CS2 courses. The goal of our project is to develop a way to provide students a library that relieves them from low-level issues of reading and parsing raw data from web-based data sources and that interfaces with data structures and representations defined by students themselves. In addition, the library requires minimal syntactic overhead to use its functionality and allows students and instructors to focus on algorithmic exercises involving processing live and large data obtained from the Internet. At a minimum, the library should serve to create drop-in replacements for traditional programming exercises in introductory courses - raising the engagement level by having students deal with "real" data rather than artificial data provided through standard input.

Proof-carrying code (PCC) is a general framework for verifying the safety properties of machine-language programs. PCC proofs are usually written in a logic extended with language-specific typing rules. In foundational proof-carrying code (FPCC), on the other hand, proofs are constructed and verified using strictly the foundations of mathematical logic, with no type-specific axioms. FPCC is more flexible and secure because it is not tied to any particular type system and it has a smaller trusted base. Foundational proofs, however are much harder to construct. Previous efforts on FPCC all required building sophisticated semantic models for types. In this paper, we present a syntactic approach to FPCC that avoids the difficulties of previous work. Under our new scheme, the foundational proof for a typed machine program simply consists of the typing derivation plus the formalized syntactic soundness proof for the underlying type system. We give a translation from a typed assembly language into FPCC and demonstrate the advantages of our new system via an implementation in the Coq proof assistant.