2018 Summer Internship Program - Openings

A wide range of internships are available for summer 2018.  This list will be updated as new internships become available, and ISI researchers also may have opportunities that aren't yet listed here. We encourage you to apply if your interests relate to any facet of ISI’s research, even if you don't see a specific project of interest here. Please see How to Apply for more information.

Jump to a broad topic area:

Research in Natural Language Understanding - Boston/Waltham Location

Natural language understanding (NLU) provides powerful technologies that help people get the information they need from spoken and written language, whether it be podcasts, newspaper articles, social media, or many other possible sources.

We seek interns to work with our Boston location's Nautral Language researchers to expand the boundaries of how machines can understand human languages. We are developing systems that are more accurate, work in more languages, use less data for training, and are able to be applied to more diverse domains. Current projects in this area include:

  • Automatic extraction and organization of knowledge from diverse information sources (news articles, social media, etc.).
  • Development of a multilingual common semantic space to enable cross-language information transfer (translation, retrieval, and summarization) for low-resource languages.
  • Developing linguistic resources for non-English languages.
  • Creation of a new open source information extraction toolkit to enable reasoning and experimentation using diverse document interpretations.

The ideal intern candidate will have experience writing Python scripts performing data manipulation, training/testing machine learning systems, and will have completed coursework in natural language processing and machine learning. Strong candidates without experience in natural language processing or machine learning are still encouraged to apply.

Location: Waltham, MA

Applications of Natural Language Understanding - Boston/Waltham Location

Natural language understanding (NLU) is providing powerful new aids that automatically pinpoint information the user needs, without requiring the user to specify in detail what they need in advance. Near-term applications include law enforcement, aids to analysts, and estimating the likelihood of geopolitical events. In one of our projects, NLU tools are being tested on web data and being used by hundreds of individuals.

Interns will work with researchers to adapt existing NLU tools, integrating them into applications, extending their functional capabilities, and making them more effective in practice.

The ideal candidate will have experience in software engineering in languages such as Python, Java, or C++, and will have completed coursework in natural language processing and/or machine learning.

Location: Waltham, MA

Intelligent Systems - Social Behavior Modeling

An individual's subjective well-being (SWB) is affected by socio-economic attributes, whcih define where the individual lives, works, shops, and who she spends time with. Even such intrinsic facets of human behavior as personality, political affiliation, health, use are affected by these demographic and socio-economic factors, which in turn affect his or her SWB. We are creating a framework for real-time monitoring and analysis of SWB and individual behavior that will lead to data-informed policies and interventions to improve well-being.  Our framework leverages social media data and extends it by incorporating additional data sources from economic, socio-cultural and educational activities, transportation, crime and other activities. 

The ideal candidate is a computationally-minded and strongly motivated student with a clear understanding of machine learning methods and applications. Python coding fluency and previous experience with social data mining are preferred.

Location: Marina del Rey, CA

Intelligent Systems - Deep Learning Approaches to Image Detection and Recognition

The computer vision and document image processing group has developed a suite of deep learning based techniques for various tasks including face recognition and OCR. Internship students will have the opportunity to apply deep learning techniques to the related problems of detecting and recognizing text in video, both overlaid text and in-scene text.  Candidates should have prior experience with Python and Python-based deep learning tools, such as Keras, Theano, Torch, or TensorFlow.

Location:  Arlington, VA

Intelligent Systems - Large-Scale Graph Analytics

This research topic focuses on scalable graph analytics techniques that are suitable for very-large scale real-world data. In particular, the project aims at developing a unified framework for heterogeneous and time-dependent network embedding, which reflects both the local and global network structures in a shared latent space, and makes the embedded results useful for a variety of big data analytic tasks. Applications areas of interest include, but are not limited to social and socio-technical network analysis, system monitoring, traffic analysis, anomaly detection, and image and video processing.

The ideal candidate will have a basic understanding of machine learning methods and applications. Java coding fluency and previous experience with graph analytics methods are preferred.

Location:  Marina del Rey, CA

Intelligent Systems - Computational Social Network Science 

Computational social sciences focus on the study of individual behaviour and society through the lens of computational and statistical tools. The selected candidate will be involved in projects related to the study of technology and society, focusing on online social networks (e.g., Twitter, Reddit), techno-social systems (e.g., mobile communication systems), news platforms, or human mobility and sensor data, by applying a combination of statistical methods, network science models, and data-driven analysis.

The ideal candidate is a computationally-minded and strongly motivated student with a clear understanding of network science and statistical modelling methods and applications. Python coding fluency and previous experience with social network mining are preferred.

Location:  Marina del Rey, CA

Intelligent Systems - Social Media Analysis: Manipulation and Abuse Dynamics

Social media have become pervasive tools for planetary scale communication and now play a central role in our society: from political discussion to social issues, from entertainment to business, such platforms shape the real-time worldwide conversation. But with new technologies, also come abuse: social media have been used for malicious activities including public opinion manipulation, propaganda, and coordination of cyber-attacks. The selected candidate will work on projects related to social media analysis, in particular studying the behaviour of social media users, and the dynamics of use and abuse of social platforms for a variety of purposes including spreading of fake news and social bots.

The ideal candidate is a computationally-minded and strongly motivated student with a clear understanding of social networks, machine learning, and data science methods and applications. Python coding fluency and previous experience with social media mining are preferred.

Location:  Marina del Rey, CA

Intelligent Systems - Machine Learning for Building Knowledge Graphs

A knowledge graph is a set of linked knowledge and one of the challenges is how to create such graphs of knowledge from existing sources. In previous work, we developed methods to map data sources to a domain ontology to describe the contents of the sources and we developed techniques to extract and use unstructured data to construct knowledge graphs. The next step in this work is to develop new and improved techniques for learning the models of sources, extracting unstructured data, and linking the data across sources.

Websites: karma.isi.edu and dig.isi.edu

Location:  Marina del Rey, CA

Intelligent Systems - Provenance for big data processing in science with Apache Spark

This project focuses on tracking provenance in scientific data analysis in SciSpark. SciSpark is a framework for scaling scientific computations that extends Apache Spark with new capabilities for processing with science data standards for large-scale data, with a particular focus on interactive climate analytics. This work is a collaboration with NASA/JPL.

More details can be found at https://github.com/SciSpark.

Location:  Marina del Rey, CA

Computing Systems - In-situ workflows for exascale computing

As a result of the growing gap between CPU and storage performance, future exascale systems will be constrained in terms of I/O. It is expected that the storage bandwidth of exascale systems will not increase significantly over existing petascale systems, while computing resources increase by orders of magnitude. Existing approaches where science applications write large, periodic checkpoint files, which are later read for analysis and visualization will not be feasible on exascale systems.  Instead, applications will need to perform data analysis and visualization in-situ as the simulation is running. In order to facilitate this new approach, tools and techniques need to be developed that enable scientists and application developers to easily compose in-situ applications. One approach is to develop in-situ workflows that enable loosely-coupled analysis and visualization components run alongside simulation components to consume data as it is generated. The development of such workflows will require new techniques for composing loosely-coupled applications, for describing the data dependencies between components, and for scheduling and optimizing data movement at runtime. This project would involve research in one of the following areas:

  • Investigate techniques for naming and transferring data between components in an in-situ workflow
  • Researching techniques for scheduling computations and data to minimize data movement in in-situ workflows
  • Developing new programming techniques and APIs for composing in-situ workflows made of existing, loosely-coupled components

Each of these areas would involve identifying a target use-case, developing a prototype solution, and evaluating the solution using a combination of physical and simulation experiments.

Location: Marina del Rey, CA

Computing Systems - Tools and Applications for Heterogeneous, Reconfigurable Computing Environments

The Reconfigurable Computing Group at USC/ISI is seeking intern applicants in the areas of application mapping to heterogeneous computing environments, CAD/EDA tool development for FPGAs, and reliability and resiliency of FPGA VLSI architectures. Interns will be actively engaged in state of the art research projects for NASA, DARPA, and IARPA, contributing to research goals such as determining the next generation space architecture for NASA Earth science missions, developing programming models and APIs which help raise the level of abstraction for heterogeneous compute elements, evaluating the trade space of mapping data analytics algorithms to FPGA architectures, enhancing the security and trust for Cybersecurity hardware, or contributing to our open source CAD tools which target real physical devices. Ideal candidates will be pursuing graduate degrees in Electrical Engineering, Computer Engineering, or Computer Science, and possess excellent skills in Verilog/VHDL/C programming for FPGAs.

Location: Arlington, VA

Computing Systems - High Performance and Real-time Cloud Computing

The USC/ISI cloud and multicore team is seeking applications for internships in the areas of high performance real-time cloud computing, cloud reliability, and real-time cloud scheduling. As an intern, you will engage in cutting-edge research at the intersection of virtualization, high performance, and reliability. You may contribute to improving cloud and application-level fail-over, or you may develop dynamic scheduling support for virtualized and heterogeneous real-time applications. The team that you will join is composed of cloud, virtualization, and real-time experts who have deep experience with OpenStack, hypervisors, accelerators, and high speed networking fabrics. Ideal candidates will have experience with Python and C/C++ and some combination of OpenStack, real-time scheduling, hypervisors, and CUDA.

Location:  Arlington, VA

Internet and Networked Systems - Software-Defined Networking Research

ISI's Internet and Networked Systems Division seeks summer interns to contribute to a project concerning distributed software-defined controllers.

Software-defined networking (SDN) out-sources the control over the network switching fabric (the data plane) to a logically centralized control plane. The advent of SDN has opened up the ability of dynamically deploying several vital applications such as traffic engineering, inter-domain routing, security and access policy to network virtualization. In many deployments the control plane ends up being actually centralized, yielding both a single point of failure and of attack. This project models the interaction between the data plane and a distributed control plane consisting of a set of failure-prone and potentially malicious control
devices, and implements a secure and robust controller platform that allows network administrators to integrate new network functionality as with a centralized approach.

An ideal candidate for this internship would have strong academic standing in algorithms, networking and operating systems, but does not necessarily require prior knowledge of SDN or distributed computing. Programming proficiency in C/C++/Python is necessary. Accepted candidates will have the chance to learn more about the internals of networked systems, cryptographic techniques and distributed algorithms.

Location: Marina del Rey, CA

Internet and Networked Systems - Tools and Methodologies for Cyber Security Experimentation

ISI’s Internet and Networked Systems Division is engaged in a unique research program targeting the creation of new methodologies for rigorous, scientifically grounded experimental cyber security research. This effort is aimed at developing new, next generation methodologies, tools, and infrastructures for use in experimental research environments, with particular application to the security of large-scale, complex networked and cyberphysical systems.

A key aspect of our work in this area is the ongoing development and operation of DETERLab, a unique cloud-like systems modeling and emulation environment established to support experimental cybersecurity research. We seek interns with interests across all aspects of experimental research methodologies, including such areas as networked systems testbed design, experimental data representation and analysis, and high-level, language-based system and process representations. Familiarity and background with experimental testbeds such as DETER, Emulab, or GENI is a plus, as is familiarity with advanced simulation methodologies, system representation tools, and/or highly concurrent programming models and environments.

Location:  Marina del Rey, CA

Internet and Networked Systems - Cyber Security Research

ISI's Internet and Networked Systems Division seeks summer interns to contribute to a number of unique, cutting-edge research projects spanning the broad space of cybersecurity and trustworthy systems. Particular topics under investigation include

  • Malware study: Safe live-malware experimentation, Detecting and engaging anti-virtualization malware.
  • Internet Measurement and Analysis: Defenses from flash-crowd DDoS attacks, DDOS Botnet detection and discrimination, Network Reconnaissance and Anomaly Detection
  • Privacy: Privacy-safe Network Trace Collection and Sharing, Anonymity in the Internet
  • Authorization, Access Control & Trust: Policy Management for Federated Systems, Software Attestation of Embedded Devices
  • Cyber-physical Systems: Smart Grid Cyber Security, Hybrid Control + Network Systems Experimentation
  • Sociology and Human Factors: Novel password design approaches, Human behavioral modeling in multi-agent systems.

We seek interns with demonstrated interest in cybersecurity research and significant familiarity with one of the above problem domains or similar; strong background knowledge of operating systems and networks; some experience with Linux, FreeBSD, or a similar Unix-derived OS; and strong programming skills in any language.

Location:  Marina del Rey, CA