@InProceedings{Rizvi23a,
        author =        "{A S M} Rizvi and Jelena Mirkovic and John
 Heidemann adn Wes Hardaker and Robert Story",
 title = "Defending Root {DNS} Servers Against {DDoS} Using Layered Defenses",
        booktitle =     "Proceedings of the " # "{IEEE} International Conference on Communications Systems and Netowrks (COMSNETS)",
        year =          2023,
	sortdate = 		"2023-01-03", 
	project = "ant, ddidd, paaddos",
	jsubject = "network_security",
	note = "Awarded best paper",
        pages =      "to appear",
        month =      jan,
        address =    "Bengaluru, India",
        publisher =  "IEEE",
        location =   "johnh: pafile",
        keywords =   "ddidd, ddos, filtering, frade",
        xdoi =        "tbd",
	blogurl = "https://ant.isi.edu/blog/?p=1948",
	url =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi23a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi23a.pdf",
	abstract = "Distributed Denial-of-Service (DDoS) attacks exhaust resources, leaving a server unavailable to legitimate clients. The Domain Name System (DNS) is a frequent target of DDoS attacks. Since DNS is a critical infrastructure service, protecting it from DoS is imperative. Many prior approaches have focused on specific filters or anti-spoofing techniques to protect generic services. DNS root nameservers are more challenging to protect, since they use fixed IP addresses, serve very diverse clients and requests, receive predominantly UDP traffic that can be spoofed, and must guarantee high quality of service. In this paper we propose a layered DDoS defense for DNS root nameservers. Our defense uses a library of defensive filters, which can be optimized for different attack types, with different levels of selectivity. We further propose a method that automatically and continuously evaluates and selects the best combination of filters throughout the attack. We show that this layered defense approach provides exceptional protection against all attack types using traces of real attacks from a DNS root nameserver. Our automated system can select the best defense within seconds and quickly reduce the traffic to the server within a manageable range while keeping collateral damage lower than 2\%. We can handle millions of filtering rules without noticeable operational overhead.",
}

Distributed Denial-of-Service (DDoS) attacks exhaust resources, leaving a server unavailable to legitimate clients. The Domain Name System (DNS) is a frequent target of DDoS attacks. Since DNS is a critical infrastructure service, protecting it from DoS is imperative. Many prior approaches have focused on specific filters or anti-spoofing techniques to protect generic services. DNS root nameservers are more challenging to protect, since they use fixed IP addresses, serve very diverse clients and requests, receive predominantly UDP traffic that can be spoofed, and must guarantee high quality of service. In this paper we propose a layered DDoS defense for DNS root nameservers. Our defense uses a library of defensive filters, which can be optimized for different attack types, with different levels of selectivity. We further propose a method that automatically and continuously evaluates and selects the best combination of filters throughout the attack. We show that this layered defense approach provides exceptional protection against all attack types using traces of real attacks from a DNS root nameserver. Our automated system can select the best defense within seconds and quickly reduce the traffic to the server within a manageable range while keeping collateral damage lower than 2%. We can handle millions of filtering rules without noticeable operational overhead.

@InProceedings{Imana23a,
        author =        "Basileal Imana and Aleksandra Korolova and John Heidemann",
        title =         "Having your Privacy Cake and Eating it Too: Platform-supported Auditing of Social Media Algorithms for Public Interest",
        booktitle =     " Computer Supported Cooperative Work ",
        year =          2023,
	sortdate = 	"2023-10-13", 
	project = "ant",
	jsubject = "network_observation",
        pages =      "to appear",
        month =      oct,
        address =    "Minneapolis, Minnesota, USA",
        publisher =  "ACM",
        keywords =   "linkedin, facebook, ad delivery algorithm, bias,
                  skew, discrimination, platform-supported auditing,
                  differential privacy",
        doi =     "xxx",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Imana23a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Imana23a.pdf",
	blogurl = "https://ant.isi.edu/blog/?p=1889",
	abstract = "Relevance estimators are algorithms used by major social media platforms to determine what content is shown to users and its presentation order. These algorithms aim to personalize the platforms' experience for users, increasing engagement and, therefore, platform revenue. However, at the large scale of many social media platforms, many have concerns that the relevance estimation and personalization algorithms are opaque and can produce outcomes that are harmful to individuals or society. Legislations have been proposed in both the U.S. and the E.U. that mandate auditing of social media algorithms by external researchers. But auditing at scale risks disclosure of users' private data and platforms' proprietary algorithms, and thus far there has been no concrete technical proposal that can provide such auditing. Our goal is to propose a new method for platform-supported auditing that can meet the goals of the proposed legislations. The first contribution of our work is to enumerate these challenges and the limitations of existing auditing methods to implement these policies at scale. Second, we suggest that limited, privileged access to relevance estimators is the key to enabling generalizable platform-supported auditing of social media platforms by external researchers. Third, we show platform-supported auditing need not risk user privacy nor disclosure of platforms' business interests by proposing an auditing framework that protects against these risks. For a particular fairness metric, we show that ensuring privacy imposes only a small constant factor increase (6.34x as an upper bound, and 4x for typical parameters) in the number of samples required for accurate auditing. Our technical contributions, combined with ongoing legal and policy efforts, can enable public oversight into how social media platforms affect individuals and society by moving past the privacy-vs-transparency hurdle."
,}

Relevance estimators are algorithms used by major social media platforms to determine what content is shown to users and its presentation order. These algorithms aim to personalize the platforms' experience for users, increasing engagement and, therefore, platform revenue. However, at the large scale of many social media platforms, many have concerns that the relevance estimation and personalization algorithms are opaque and can produce outcomes that are harmful to individuals or society. Legislations have been proposed in both the U.S. and the E.U. that mandate auditing of social media algorithms by external researchers. But auditing at scale risks disclosure of users' private data and platforms' proprietary algorithms, and thus far there has been no concrete technical proposal that can provide such auditing. Our goal is to propose a new method for platform-supported auditing that can meet the goals of the proposed legislations. The first contribution of our work is to enumerate these challenges and the limitations of existing auditing methods to implement these policies at scale. Second, we suggest that limited, privileged access to relevance estimators is the key to enabling generalizable platform-supported auditing of social media platforms by external researchers. Third, we show platform-supported auditing need not risk user privacy nor disclosure of platforms' business interests by proposing an auditing framework that protects against these risks. For a particular fairness metric, we show that ensuring privacy imposes only a small constant factor increase (6.34x as an upper bound, and 4x for typical parameters) in the number of samples required for accurate auditing. Our technical contributions, combined with ongoing legal and policy efforts, can enable public oversight into how social media platforms affect individuals and society by moving past the privacy-vs-transparency hurdle.

@inproceedings{schmidt2023_1,
	author = {Andrew G. Schmidt and Justinand Benedict Reynwar and Kellie Canida and Matthew French},
	booktitle = {Government Microcircuit Applications \& Critical Techology Conference (GOMACTech)},
	title = {Integrity and Assurance Tools Targeting FPGA Hard IP Block Bitstream Undocumented Functionality},
	year = {2023}}

@inproceedings{schmidt2023_3,
	author = {D. Soni and N. Neda and N. Zhang and Benedict Reynwar and B. Heyman and M. Moopan and A. Badawi and Y. Polyakov and Kellie Canidaand Anrew G. Schmidt and M. Pedram and D. Cousins and Matthew French and F. Franchetti and R. Karri and M. Maniatakos and B. Reagen},
	title = {RPU: A Ring Processing Unit with Applications to FHE},
	booktitle = {IEEE International Symposium on Performance Analysis of Systems and Software (ISPAS2023)},
	year = {2023}}

@inproceedings{schmidt2023_2,
	author = {D. Cousins and Y. Polyakov and A. Badawi and Andrew G. Schmidt and Justinand Benedict Reynwar and Kellie Canida and Matthew French and A.Jaiswal and C. Mathew and H. Gamil and N. Neda and D. Soni and M. Maniatakos and B.Reagen},
	booktitle = {Government Microcircuit Applications \& Critical Techology Conference (GOMACTech)},
	title = {TREBUCHET: Fully Homomorphic Encryption Accelerator for Deep Computation},
	year = {2023}}

@article{dibaji2022tutorial,
  title={A Tutorial on Security and Privacy Challenges in CPS},
  author={Dibaji, Seyed Mehran and Hussain, Alefiya and Ishii, Hideaki},
  journal={Security and Resilience of Control Systems},
  pages={121--146},
  year={2022},
  publisher={Springer, Cham}
}

@InProceedings{Rizvi22a,
        author =        "{A S M} Rizvi and Leandro Bertholdo and
 Jo{\~a}o Ceron and John Heidemann",
 title = "Anycast Agility: Network Playbooks to Fight {DDoS}",
        booktitle =     "Proceedings of the " # "31st" # " {USENIX} Security Symposium ",
        year =          2022,
	sortdate = 		"2022-08-10", 
	project = "ant, ddidd, paaddos, sabres",
	jsubject = "network_security",
        pages =      "4201--4218",
        month =      aug,
        publisher =  "{USENIX}",
        location =   "johnh: pafile",
        keywords =   "anycast, ddos, routing",
        doi =        "to appear",
	blogurl = "https://ant.isi.edu/blog/?p=1829",
	url =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi22a.pdf",
	oldotherurl = "https://www.usenix.org/system/files/sec22fall_rizvi.pdf",
	otherurl = "https://www.usenix.org/conference/usenixsecurity22/presentation/rizvi",
	dataurl =	"https://zenodo.org/record/6473023",
	talkurl = "https://www.usenix.org/conference/usenixsecurity22/presentation/rizvi",
	abstract = "IP anycast is used for services such as DNS and Content Delivery
Networks (CDN) to provide the capacity to handle Distributed
Denial-of-Service (DDoS) attacks.  During a DDoS attack service
operators redistribute traffic between anycast sites to take advantage
of sites with unused or greater capacity.  Depending on site traffic
and attack size, operators may instead concentrate attackers in a few
sites to preserve operation in others.  Operators use these actions
during attacks, but how to do so has not been described systematically
or publicly.  This paper describes several methods to use BGP to shift
traffic when under DDoS, and shows that a \emph{response playbook} can
provide a menu of responses that are options during an attack.  To
choose an appropriate response from this playbook, we also describe a
new method to estimate true attack size, even though the operator's
view during the attack is incomplete.  Finally, operator choices are
constrained by distributed routing policies, and not all are helpful.
We explore how specific anycast deployment can constrain options in
this playbook, and are the first to measure how generally applicable
they are across multiple anycast networks.
",
}

IP anycast is used for services such as DNS and Content Delivery Networks (CDN) to provide the capacity to handle Distributed Denial-of-Service (DDoS) attacks. During a DDoS attack service operators redistribute traffic between anycast sites to take advantage of sites with unused or greater capacity. Depending on site traffic and attack size, operators may instead concentrate attackers in a few sites to preserve operation in others. Operators use these actions during attacks, but how to do so has not been described systematically or publicly. This paper describes several methods to use BGP to shift traffic when under DDoS, and shows that a \emphresponse playbook can provide a menu of responses that are options during an attack. To choose an appropriate response from this playbook, we also describe a new method to estimate true attack size, even though the operator's view during the attack is incomplete. Finally, operator choices are constrained by distributed routing policies, and not all are helpful. We explore how specific anycast deployment can constrain options in this playbook, and are the first to measure how generally applicable they are across multiple anycast networks.

@conference {Schmidt2022,
    title = {Bitstream Assurance Checking Engine for Undocumented Functionality},
    booktitle = {Government Microcircuit Applications and Critical Technology Conference (GOMACTech)},
    year = {2022},
    month = {March},
    author = {A. Schmidt, J. Wilford, B. Reynwar, T. Sung, and M. French}
    }

    

@inproceedings{schmidt2022_0,
	author = {Andrew G. Schmidt and Justin Wilford and Benedict Reynwar and Ting-Yuan Sung and Matthew French},
	booktitle = {Government Microcircuit Applications \& Critical Techology Conference (GOMACTech)},
	title = {Bitstream Assurance Checking Engine for Undocumented Functionality (BRACE)},
	year = {2022}}

@Article{chen2022charting,
    author={Chen, Emily and Jiang, Julie and Chang, Ho-Chun Herbert and Muric, Goran and Ferrara, Emilio},
    title={Charting the Information and Misinformation Landscape to Characterize Misinfodemics on Social Media: {COVID}-19 Infodemiology Study at a Planetary Scale},
    journal={JMIR Infodemiology},
    year={2022},
    month={Feb},
    day={8},
    volume={2},
    number={1},
    pages={e32378},
    issn={2564-1891},
    doi={10.2196/32378},
}

@InProceedings{Rizvi22b,
        author =        "{A S M} Rizvi and John Heidemann",
        title =         "Chhoyhopper: A Moving Target Defense with {IPv6}",
        booktitle =     "Proceedings of the " # "{IEEE} Workshop on Measurements, Attacks, and Defenses for the Web (MADWeb)",
        year =          2022,
	sortdate = 		"2022-04-24", 
	project = "ant, ddidd, paaddos, sabres",
	jsubject = "network_security",
        pages =      "to appear",
        month =      apr,
        address =    "San Diego, California, USA",
        publisher =  "IEEE",
        location =   "johnh: pafile",
        keywords =   "chhoyhopper, moving target defense, ipv6, https, tls, ssh",
        doi =        "https://dx.doi.org/10.14722/madweb.2022.23004",
	blogurl = "https://ant.isi.edu/blog/?p=1845",
	url =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Rizvi22b.pdf",
	abstract = "Services on the public Internet are frequently scanned, then subject
to brute-force password attempts and Denial-of-Service (DoS) attacks.
We would like to run such services stealthily, where they are
available to friends but hidden from adversaries.  In this work, we
propose a discovery-resistant moving target defense named
``Chhoyhopper'' that utilizes the vast IPv6 address space to conceal
publicly available services.  The client meets the server at an IPv6
address that changes in a pattern based on a shared, pre-distributed
secret and the time of day.  By hopping over a /64 prefix, services
cannot be found by active scanners, and passively observed information
is useless after two minutes.  We demonstrate our system with the two
important applications---SSH and HTTPS, and make our system publicly
available.",
}

Services on the public Internet are frequently scanned, then subject to brute-force password attempts and Denial-of-Service (DoS) attacks. We would like to run such services stealthily, where they are available to friends but hidden from adversaries. In this work, we propose a discovery-resistant moving target defense named ``Chhoyhopper'' that utilizes the vast IPv6 address space to conceal publicly available services. The client meets the server at an IPv6 address that changes in a pattern based on a shared, pre-distributed secret and the time of day. By hopping over a /64 prefix, services cannot be found by active scanners, and passively observed information is useless after two minutes. We demonstrate our system with the two important applications—SSH and HTTPS, and make our system publicly available.

@TechReport{Moura22b,
	  author = "G. Moura and W. Hardaker and J. Heidemann and M. Davids",
	  title = "Considerations for Large Authoritative {DNS} Server Operators",
	  institution =  "Internet Request For Comments",
	  year = 	       2022,
	  type =	       "RFC",
	  number =       9199,
	  month =	       mar,
	note = "Informational",
	  location =	"johnh: pafiles",
	  keywords =     "internet routing anycast, IP anycast, dns anycast,
                  rfc-9199, root dns anycast (but not explicitly)",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Moura22b.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Moura22b.pdf",
	  otherurl =	       "https://www.rfc-editor.org/rfc/rfc9199.txt",
	doi = "http://dx.doi.org/10.17487/RFC9199",
	abstract = "
   Recent research work has explored the deployment characteristics and
   configuration of the Domain Name System (DNS).  This document
   summarizes the conclusions from these research efforts and offers
   specific, tangible considerations or advice to authoritative DNS
   server operators.  Authoritative server operators may wish to follow
   these considerations to improve their DNS services. \newline
   It is possible that the results presented in this document could be
   applicable in a wider context than just the DNS protocol, as some of
   the results may generically apply to any stateless/short-duration
   anycasted service. \newline
   This document is not an IETF consensus document: it is published for
   informational purposes.
",
}

Recent research work has explored the deployment characteristics and configuration of the Domain Name System (DNS). This document summarizes the conclusions from these research efforts and offers specific, tangible considerations or advice to authoritative DNS server operators. Authoritative server operators may wish to follow these considerations to improve their DNS services. \newline It is possible that the results presented in this document could be applicable in a wider context than just the DNS protocol, as some of the results may generically apply to any stateless/short-duration anycasted service. \newline This document is not an IETF consensus document: it is published for informational purposes.

@article{wang2022contextualized,
  title={Contextualized Scene Imagination for Generative Commonsense Reasoning},
  author={Wang, PeiFeng and Zamora, Jonathan and Liu, Junfeng and Ilievski, Filip and Chen, Muhao and Ren, Xiang},
  journal={ICLR},
  year={2022}
}

@article{wroclawski:2022uy,
	author = {John Wroclawski and Terry Benzel},
	date-added = {2022-03-14 13:56:10 -0700},
	date-modified = {2022-03-14 13:57:02 -0700},
	journal = {Communications of the ACM},
	month = {May},
	title = {Cybersecurity as Illuminator for the Future of Computing Research},
	year = {2022}}

@InProceedings{Saluja22a,
        author =        "Tarang Saluja and Johnh Heidemann and Yuri Pradkin",
        title =         "Differences in Monitoring the {DNS} Root Over {IPv4} and {IPv6}",
        booktitle =     "Proceedings of the " # " National Symposium for NSF REU Research in Data Science, Systems, and Security ",
        year =          2022,
	sortdate = 	"2022-12-15", 
	project = "ant, eieio, reu, isireu",
	jsubject = "topology_modeling",
        pages =      "to appear",
        month =      dec,
        address =    "Portland, OR, USA",
        publisher =  "IEEE",
        location =   "johnh: pafile",
        keywords =   "root server system, dnsmon, ripe atlas, ipv4, ipv6",
	doi = "xxx",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Saluja22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Saluja22a.pdf",
	dataurl = "https://ant.isi.edu/ripe_atlas_islands/",
	blogurl = "https://ant.isi.edu/blog/?p=1937",
	abstract = "The Domain Name System (DNS) is an essential service for the Internet
which maps host names to IP addresses.  The DNS Root Sever System
operates the top of this namespace.  RIPE Atlas observes DNS from more
than 11k vantage points (VPs) around the world, reporting the
reliability of the DNS Root Server System in DNSmon.  DNSmon shows
that loss rates for queries to the DNS Root are nearly 10\% for IPv6,
much higher than the approximately 2\% loss seen for IPv4.  Although
IPv6 is ``new,'' as an operational protocol available to a third of
Internet users, it ought to be just as reliable as IPv4.  We examine
this difference at a finer granularity by investigating loss at
individual VPs.  We confirm that specific VPs are the source of this
difference and identify two root causes:  VP \emph{islands} with
routing problems at the edge which leave them unable to access IPv6
outside their LAN, and VP \emph{peninsulas} which indicate routing
problems in the core of the network.  These problems account for most
of the loss and nearly all of the difference between IPv4 and IPv6
query loss rates.  Islands account for most of the loss (half of IPv4
failures and 5/6ths of IPv6 failures), and we suggest these
measurement devices should be filtered out to get a more accurate
picture of loss rates.  Peninsulas account for the main differences
between root identifiers, suggesting routing disagreements root
operators need to address.  We believe that filtering out both of
these known problems provides a better measure of underlying network
anomalies and loss and will result in more actionable alerts.",
}

The Domain Name System (DNS) is an essential service for the Internet which maps host names to IP addresses. The DNS Root Sever System operates the top of this namespace. RIPE Atlas observes DNS from more than 11k vantage points (VPs) around the world, reporting the reliability of the DNS Root Server System in DNSmon. DNSmon shows that loss rates for queries to the DNS Root are nearly 10% for IPv6, much higher than the approximately 2% loss seen for IPv4. Although IPv6 is ``new,'' as an operational protocol available to a third of Internet users, it ought to be just as reliable as IPv4. We examine this difference at a finer granularity by investigating loss at individual VPs. We confirm that specific VPs are the source of this difference and identify two root causes: VP \emphislands with routing problems at the edge which leave them unable to access IPv6 outside their LAN, and VP \emphpeninsulas which indicate routing problems in the core of the network. These problems account for most of the loss and nearly all of the difference between IPv4 and IPv6 query loss rates. Islands account for most of the loss (half of IPv4 failures and 5/6ths of IPv6 failures), and we suggest these measurement devices should be filtered out to get a more accurate picture of loss rates. Peninsulas account for the main differences between root identifiers, suggesting routing disagreements root operators need to address. We believe that filtering out both of these known problems provides a better measure of underlying network anomalies and loss and will result in more actionable alerts.

@article{oltramari2022generalizable,
  title={Generalizable Neuro-symbolic Systems for Commonsense Question Answering},
  author={Oltramari, Alessandro and Francis, Jonathan and Ilievski, Filip and Ma, Kaixin and Mirzaee, Roshanak},
  journal={arXiv preprint arXiv:2201.06230},
  year={2022}
}

@TechReport{Imana22a,
        author =        "Basileal Imana and Aleksandra Korolova and John Heidemann",
        title =         "Having your Privacy Cake and Eating it Too: Platform-supported Auditing of Social Media Algorithms for Public Interest",
        institution =   "arXiv",
        year =          2022,
	sortdate = 	"2022-07-18", 
	project = "ant",
	jsubject = "network_observation",
        number =     "arXiv:2207.08773v1 [cs.CY]",
        month =      jul,
        location =   "johnh: pafile",
        keywords =   "linkedin, facebook, ad delivery algorithm, bias,
                  skew, discrimination, platform-supported auditing,
                  differential privacy",
        doi =     "https://doi.org/10.48550/arXiv.2207.08773",
        otherurl =        "https://arxiv.org/pdf/2207.08773v1.pdf",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Imana22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Imana22a.pdf",
	blogurl = "https://ant.isi.edu/blog/?p=1889",
	abstract = "
Relevance estimators are algorithms used by major social media platforms to determine what content is shown to users and its presentation order. These algorithms aim to personalize the platforms' experience for users, increasing engagement and, therefore, platform revenue. However, at the large scale of many social media platforms, many have concerns that the relevance estimation and personalization algorithms are opaque and can produce outcomes that are harmful to individuals or society. Legislations have been proposed in both the U.S. and the E.U. that mandate auditing of social media algorithms by external researchers. But auditing at scale risks disclosure of users' private data and platforms' proprietary algorithms, and thus far there has been no concrete technical proposal that can provide such auditing. Our goal is to propose a new method for platform-supported auditing that can meet the goals of the proposed legislations. The first contribution of our work is to enumerate these challenges and the limitations of existing auditing methods to implement these policies at scale. Second, we suggest that limited, privileged access to relevance estimators is the key to enabling generalizable platform-supported auditing of social media platforms by external researchers. Third, we show platform-supported auditing need not risk user privacy nor disclosure of platforms' business interests by proposing an auditing framework that protects against these risks. For a particular fairness metric, we show that ensuring privacy imposes only a small constant factor increase (6.34x as an upper bound, and 4x for typical parameters) in the number of samples required for accurate auditing. Our technical contributions, combined with ongoing legal and policy efforts, can enable public oversight into how social media platforms affect individuals and society by moving past the privacy-vs-transparency hurdle."
}

Relevance estimators are algorithms used by major social media platforms to determine what content is shown to users and its presentation order. These algorithms aim to personalize the platforms' experience for users, increasing engagement and, therefore, platform revenue. However, at the large scale of many social media platforms, many have concerns that the relevance estimation and personalization algorithms are opaque and can produce outcomes that are harmful to individuals or society. Legislations have been proposed in both the U.S. and the E.U. that mandate auditing of social media algorithms by external researchers. But auditing at scale risks disclosure of users' private data and platforms' proprietary algorithms, and thus far there has been no concrete technical proposal that can provide such auditing. Our goal is to propose a new method for platform-supported auditing that can meet the goals of the proposed legislations. The first contribution of our work is to enumerate these challenges and the limitations of existing auditing methods to implement these policies at scale. Second, we suggest that limited, privileged access to relevance estimators is the key to enabling generalizable platform-supported auditing of social media platforms by external researchers. Third, we show platform-supported auditing need not risk user privacy nor disclosure of platforms' business interests by proposing an auditing framework that protects against these risks. For a particular fairness metric, we show that ensuring privacy imposes only a small constant factor increase (6.34x as an upper bound, and 4x for typical parameters) in the number of samples required for accurate auditing. Our technical contributions, combined with ongoing legal and policy efforts, can enable public oversight into how social media platforms affect individuals and society by moving past the privacy-vs-transparency hurdle.

@TechReport{Enayet22b,
        author =        "Asma Enayet and John Heidemann",
        title =         "Internet Outage Detection using Passive
                  Analysis (poster abstract and poster)",
        institution =   "USC/Information Sciences Institute",
        year =          "2022",
	month = sep,
	sortdate = 		"2022-09-28", 
	project = "ant, eieio",
	jsubject = "routing",
        pages =      "to appear",
        address =    "Nice, France",
        publisher =  "ACM",
        location =   "johnh: pafile",
        keywords =   "outage detection, passive data, b-root, ipv6",
        number =     "arxiv:2209.13767v3",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Enayet22b.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Enayet22b.pdf",
        doi =        "https://doi.org/10.48550/arXiv.2209.13767",
	abstract = "Outages from natural disasters, political events, software or hardware issues, and human error place a huge cost on e-commerce (\$66k per minute at Amazon). While several existing systems detect Internet outages, these systems are often too inflexible, with fixed parameters across the whole internet with CUSUM-like change detection. We instead propose a system using passive data, to cover both IPv4 and IPv6, customizing parameters for each block to optimize the performance of our Bayesian inference model. Our poster describes our three contributions: First, we show how customizing parameters allows us often to detect outages that are at both fine timescales (5 minutes) and fine spatial resolutions (/24 IPv4 and /48 IPv6 blocks). Our second contribution is to show that, by tuning parameters differently for different blocks, we can scale back temporal precision to cover more challenging blocks. Finally, we show our approach extends to IPv6 and provides the first reports of IPv6 outages."
}

Outages from natural disasters, political events, software or hardware issues, and human error place a huge cost on e-commerce ($66k per minute at Amazon). While several existing systems detect Internet outages, these systems are often too inflexible, with fixed parameters across the whole internet with CUSUM-like change detection. We instead propose a system using passive data, to cover both IPv4 and IPv6, customizing parameters for each block to optimize the performance of our Bayesian inference model. Our poster describes our three contributions: First, we show how customizing parameters allows us often to detect outages that are at both fine timescales (5 minutes) and fine spatial resolutions (/24 IPv4 and /48 IPv6 blocks). Our second contribution is to show that, by tuning parameters differently for different blocks, we can scale back temporal precision to cover more challenging blocks. Finally, we show our approach extends to IPv6 and provides the first reports of IPv6 outages.

@InProceedings{Enayet22a,
        author =        "Asma Enayet and John Heidemann",
        title =         "Internet Outage Detection using Passive
                  Analysis (poster abstract)",
        booktitle =     "Proceedings of the " # "ACM Internet Measurement Conference",
        year =          2022,
	sortdate = 		"2022-10-25", 
	project = "ant, eieio",
	jsubject = "routing",
	month = oct,
        pages =      "772--773",
        address =    "Nice, France",
        publisher =  "ACM",
        location =   "johnh: pafile",
        keywords =   "outage detection, passive data, b-root, ipv6",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Enayet22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Enayet22a.pdf",
        doi =        "https://doi.org/10.1145/3517745.3563032",
	abstract = "Outages from natural disasters, political events, software or hardware issues, and human error place a huge cost on e-commerce (\$66k per minute at Amazon). While several existing systems detect Internet outages, these systems are often too inflexible, with fixed parameters across the whole internet with CUSUM-like change detection. We instead propose a system using passive data, to cover both IPv4 and IPv6, customizing parameters for each block to optimize the performance of our Bayesian inference model. Our poster describes our three contributions: First, we show how customizing parameters allows us often to detect outages that are at both fine timescales (5 minutes) and fine spatial resolutions (/24 IPv4 and /48 IPv6 blocks). Our second contribution is to show that, by tuning parameters differently for different blocks, we can scale back temporal precision to cover more challenging blocks. Finally, we show our approach extends to IPv6 and provides the first reports of IPv6 outages."
}

Outages from natural disasters, political events, software or hardware issues, and human error place a huge cost on e-commerce ($66k per minute at Amazon). While several existing systems detect Internet outages, these systems are often too inflexible, with fixed parameters across the whole internet with CUSUM-like change detection. We instead propose a system using passive data, to cover both IPv4 and IPv6, customizing parameters for each block to optimize the performance of our Bayesian inference model. Our poster describes our three contributions: First, we show how customizing parameters allows us often to detect outages that are at both fine timescales (5 minutes) and fine spatial resolutions (/24 IPv4 and /48 IPv6 blocks). Our second contribution is to show that, by tuning parameters differently for different blocks, we can scale back temporal precision to cover more challenging blocks. Finally, we show our approach extends to IPv6 and provides the first reports of IPv6 outages.

@inproceedings{chami2022taxonomy,
  author={Ines Chami AND Sami Abu-El-Haija AND Bryan Perozzi AND Christopher Re AND Kevin Murphy},
  title={Machine Learning on Graphs: A Model and Comprehensive Taxonomy},
  booktitle={Journal of Machine Learning Research},
  year={2022},
}

@InProceedings{Moura22a,
        author =        "Giovane C. M. Moura and John Heidemann and
 Wes Hardaker and Pithayuth Charnsethikul and Jeroen
 Bulten and Jo{\~a}o M. Ceron and Cristian Hesselman",
 title = "Old but Gold: Prospecting {TCP} to Engineer and Live  Monitor {DNS} Anycast",
        booktitle =     "Proceedings of the " # " Passive and Active Measurement Workshop",
	project = "ant, paaddos, ddidd",
	jsubject = "network_security",
        year =          2022,
	sortdate = 	"2022-03-28",
        pages =      "to appear",
        month =      mar,
	note = "best paper award",
        address =    "virtual",
        publisher =  "Springer",
        location =   "johnh: pafile",
	keywords = 	"anycast, dns, tcp, latency, root, .nl-tld, monitoring",
        doi =        "tbd",
	url =		"https://www.isi.edu/%7ejohnh/PAPERS/Moura22a.html",
	pdfurl =	"https://www.isi.edu/%7ejohnh/PAPERS/Moura22a.pdf",
	blogurl = "https://ant.isi.edu/blog/?p=1854",
	abstract = "DNS latency is a concern for many service operators:  CDNs exist to
reduce service latency to end-users but must rely on global DNS for
reachability and load-balancing.  Today, DNS latency is monitored by
active probing from distributed platforms like RIPE Atlas, with
Verfploeter, or with commercial services.  While Atlas coverage is
wide, its 10k sites see only a fraction of the Internet.  In this
paper we show that passive observation of TCP handshakes can measure
\emph{live DNS latency, continuously, providing good coverage of
current clients of the service}.  Estimating RTT from TCP is an old
idea, but its application to DNS has not previously been studied
carefully.  We show that there is sufficient TCP DNS traffic today to
provide good operational coverage (particularly of IPv6), and very
good temporal coverage (better than existing approaches), enabling
near-real time evaluation of DNS latency from \emph{real clients}.  We
also show that DNS servers can optionally solicit TCP to broaden
coverage.  We quantify coverage and show that estimates of DNS latency
from TCP is consistent with UDP latency.  Our approach finds
previously unknown, real problems:  \emph{DNS polarization} is a new
problem where a hypergiant sends global traffic to one anycast site
rather than taking advantage of the global anycast deployment.
Correcting polarization in Google DNS cut its latency from 100ms to
10ms; and from Microsoft Azure cut latency from 90ms to 20ms.
We also show other instances of routing problems that add 100--200ms
latency.  Finally, \emph{real-time} use of our approach for a European
country-level domain has helped detect and correct a BGP routing
misconfiguration that detoured European traffic to Australia.  We have
integrated our approach into several open source tools:  Entrada, our
open source data warehouse for DNS, a monitoring tool (ANTS), which
has been operational for the last 2 years on a country-level top-level
domain, and a DNS anonymization tool in use at a root server since
March 2021.",
}

DNS latency is a concern for many service operators: CDNs exist to reduce service latency to end-users but must rely on global DNS for reachability and load-balancing. Today, DNS latency is monitored by active probing from distributed platforms like RIPE Atlas, with Verfploeter, or with commercial services. While Atlas coverage is wide, its 10k sites see only a fraction of the Internet. In this paper we show that passive observation of TCP handshakes can measure \emphlive DNS latency, continuously, providing good coverage of current clients of the service. Estimating RTT from TCP is an old idea, but its application to DNS has not previously been studied carefully. We show that there is sufficient TCP DNS traffic today to provide good operational coverage (particularly of IPv6), and very good temporal coverage (better than existing approaches), enabling near-real time evaluation of DNS latency from \emphreal clients. We also show that DNS servers can optionally solicit TCP to broaden coverage. We quantify coverage and show that estimates of DNS latency from TCP is consistent with UDP latency. Our approach finds previously unknown, real problems: \emphDNS polarization is a new problem where a hypergiant sends global traffic to one anycast site rather than taking advantage of the global anycast deployment. Correcting polarization in Google DNS cut its latency from 100ms to 10ms; and from Microsoft Azure cut latency from 90ms to 20ms. We also show other instances of routing problems that add 100–200ms latency. Finally, \emphreal-time use of our approach for a European country-level domain has helped detect and correct a BGP routing misconfiguration that detoured European traffic to Australia. We have integrated our approach into several open source tools: Entrada, our open source data warehouse for DNS, a monitoring tool (ANTS), which has been operational for the last 2 years on a country-level top-level domain, and a DNS anonymization tool in use at a root server since March 2021.

@inproceedings{schmidt2022_5,
	author = {G. Datta, S. Kundu, Z. Yin, J. Mathai, Z. Liu, Z. Wang, M. Tian, S. Lu, R. T. Lakkireddy, Andrew G. Schmidt, W. Abd-Almageed, A. P. Jacob, A. R Jaiswal, P. A. Beerel},
	title = {P2M-DeTrack: Processing-in-Pixel-in-Memory for Energy-efficient and Real-Time Multi-Object Detection and Tracking},
	booktitle = {IFIP IEEE 30th International Conference on Very Large Scale Integration (VLSI SoC)},
	year = {2022}}

@inproceedings{polydetect,
  author={Xiyue Deng and Jelena Mirkovic},
  title={Polymorphic Malware Behavior Through Network Trace Analysis},
  booktitle={Proceedings of 14th International Conference on COMmunication Systems & NETworkS (COMSNETS)},
  year={2022}
}

@conference {Carr2022,
    title = {StereoBit on the SpaceCube Mini},
    booktitle = {IEEE International Geoscience and Remote Sensing Symposium},
    year = {2022},
    author = {J. Carr, C. Wilson, D. Wu, M. French, M. Paolieri, H. Madani, M. Kelly}
}

@inproceedings{schmidt2023_4,
	author = {N. Zhang and H. Gamil and P. Brinich and B. Reynwar and A. Badawi and N. Neda and D. Soni and K. Canida and Y. Polyakov and P. Broderick and M. Maniatakos and Andrew G. Schmidt and Matthew Franch and J. Johnson and B. Reagen and D. Bruce Cousins and F. Franchetti},
	title = {Towards Full-Stack Acceleration for Fully Homomorphic Encryption},
	booktitle = {IEEE High Performance Extreme Computing Conference (HPEC)},
	year = {2022}}

@article{kejriwal2022transfer,
  title={Transfer-based taxonomy induction over concept labels},
  author={Kejriwal, Mayank and Shen, Ke and Ni, Chien-Chun and Torzec, Nicolas},
  journal={Engineering Applications of Artificial Intelligence},
  volume={108},
  pages={104548},
  year={2022},
  month={2},
  publisher={Elsevier}
}

@article{burghardt2022unequal,
  title={Unequal impact and spatial aggregation distort covid-19 growth rates},
  author={Burghardt, Keith and Guo, Siyi and Lerman, Kristina},
  journal={Philosophical Transactions of the Royal Society A},
  volume={380},
  number={2214},
  pages={20210122},
  year={2022},
  publisher={The Royal Society}
}

@conference {Chen2022,
    title = {Untangling IP Protection via Learning and Structure},
    booktitle = {Government Microcircuit Applications and Critical Technology Conference (GOMACTech)},
    year = {2022},
    month = {March},
    author = {D. Chen, X. Zhou, S. Chowdhury, P. Beerel, P. Nuzzo, and M. French}
    }

@article{nghiem2021stop,
  title={" Stop Asian Hate!": Refining Detection of Anti-Asian Hate Speech During the COVID-19 Pandemic},
  author={Nghiem, Huy and Morstatter, Fred},
  journal={arXiv preprint arXiv:2112.02265},
  year={2021}
}

@article{spangher2021don,
  title={"Don't quote me on that": Finding Mixtures of Sources in News Articles},
  author={Spangher, Alexander and Peng, Nanyun and May, Jonathan and Ferrara, Emilio},
  journal={arXiv preprint arXiv:2104.09656},
  year={2021}
}

@article{chen2021election2020,
  title={\#Election2020: The First Public Twitter Dataset on the 2020 US Presidential Election},
  author={Chen, Emily and Deb, Ashok and Ferrara, Emilio},
  journal={Journal of Computational Social Science},
  year={2021}
}

@article{chang2021justiceforgeorgefloyd,
  title={\#JusticeForGeorgeFloyd: How Instagram Facilitated the 2020 Black Lives Matter Protests},
  author={Chang, Ho-Chun Herbert and Richardson, Allissa and Ferrara, Emilio},
  year={2021},
  publisher={SocArXiv}
}

@Misc{		  scitech2021symposium,
  Author	= {Hayes, Cassandra and Kulkarni, Chaitra and Milman, Eric D.
		  and Okunloye, Oluwabusayo and Olshansky, Alex and Oruche,
		  Roland and Kee, Kerk and Moreira, Priscila C. S. and
		  Vardeman, Charles and Coleman, Tain\=a and Do, Tu Mai Anh
		  and Jain, Aditi and Krawczuk, Patrycja and Lam, Kelsie and
		  Nagarkar, Shubham and Papadimitriou, George and Subramanya,
		  Srujana and White, Rebecca and Whitcup, Wendy and Ferreira
		  da Silva, Rafael and Deelman, Ewa},
  Title		= {{2021 SciTech and Friends Research Symposium}},
  Month		= {May},
  Year		= {2021},
  Publisher	= {Zenodo},
  DOI		= {10.5281/zenodo.4847543},
  URL		= {https://doi.org/10.5281/zenodo.4847543}
}

@ARTICLE{2021arXiv210308464K,
       author = {{Kowalsky}, Matthew and {Albash}, Tameem and {Hen}, Itay and {Lidar}, Daniel A.},
        title = "{3-Regular 3-XORSAT Planted Solutions Benchmark of Classical and Quantum Heuristic Optimizers}",
      journal = {arXiv e-prints},
     keywords = {Quantum Physics},
         year = 2021,
        month = mar,
          eid = {arXiv:2103.08464},
        pages = {arXiv:2103.08464},
archivePrefix = {arXiv},
       eprint = {2103.08464},
 primaryClass = {quant-ph},
       adsurl = {https://ui.adsabs.harvard.edu/abs/2021arXiv210308464K},
      adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}

@InProceedings{dhinagar2021:sipaim,
  author = 	 {Nikhil J. Dhinagar and Sophia I. Thomopoulos and Conor Owens-Walton and Dimitris Stripelis and Jos\'{e} Luis Ambite and Greg Ver Steeg and Daniel Weintraub and Philip Cook and Corey McMillan and Paul M. Thompson},
  title = 	 {3D Convolutional Neural Networks for Classification of Alzheimer's and Parkinson's Disease with T1-Weighted Brain MRI},
  booktitle = {17th International Symposium on Medical Information Processing and Analysis {(SIPAIM)}},
  year = 	 {2021},
  address = 	 {Campinas, Brazil},
  ISIArea = {KT,ML},

}

@inproceedings{nikhil,
	Author = {Nikhil J. Dhinagar and Sophia I. Thomopoulos and Conor Owens-Walton and Dimitris Stripelis and Jose Luis Ambite and Greg {Ver Steeg} and Daniel Weintraub and Philip Cook and Corey McMillan and Paul M. Thompson},
	Booktitle = {International Symposium on Medical Information Processing and Analysis (SIPAIM)},
	Date-Added = {2021-09-01 16:04:24 -0700},
	Date-Modified = {2021-09-01 16:27:50 -0700},
	Title = {3D Convolutional Neural Networks for Classification of Alzheimer's and Parkinson's Disease with T1-Weighted Brain MRI},
	Year = {2021}}

@InProceedings{	  ketron-escience-2021,
  Author	= {Ketron, R. and Leonard, J. and Roachell, B. and Patel, R.
		  and White, R. and Caíno-Lores, S. and Tan, N. and Miles,
		  P. and Vahi, K. and Deelman, E. and Brown, D. and Taufer,
		  M.},
  BookTitle	= {2021 IEEE 17th International Conference on eScience
		  (eScience)},
  Title		= {A Case Study in Scientific Reproducibility from the Event
		  Horizon Telescope (EHT)},
  Year		= {2021},
  Volume	= {},
  Number	= {},
  Pages		= {249-250},
  DOI		= {10.1109/eScience51609.2021.00045}
}

@InProceedings{	  community2021works,
  Title		= {A Community Roadmap for Scientific Workflows Research and
		  Development},
  Author	= {Ferreira da Silva, Rafael and Casanova, Henri and Chard,
		  Kyle and Altintas, Ilkay and Badia, Rosa M and Balis,
		  Bartosz and Coleman, Tain\~a and Coppens, Frederik and Di
		  Natale, Frank and Enders, Bjoern and Fahringer, Thomas and
		  Filgueira, Rosa and Fursin, Grigori and Garijo, Daniel and
		  Goble, Carole and Howell, Dorran and Jha, Shantenu and
		  Katz, Daniel S. and Laney, Daniel and Leser, Ulf and
		  Malawski, Maciej and Mehta, Kshitij and Pottier, Lo\"ic and
		  Ozik, Jonathan and Peterson, J. Luc and Ramakrishnan,
		  Lavanya and Soiland-Reyes, Stian and Thain, Douglas and
		  Wolf, Matthew},
  BookTitle	= {2021 IEEE Workshop on Workflows in Support of Large-Scale
		  Science (WORKS)},
  Pages		= {81--90},
  Year		= {2021},
  DOI		= {10.1109/WORKS54523.2021.00016}
}

@article{nettasinghe2021directed,
  title={A Directed, Bi-Populated Preferential Attachment Model with Applications to Analyzing the Glass Ceiling Effect},
  author={Nettasinghe, Buddhika and Alipourfard, Nazanin and Krishnamurthy, Vikram and Lerman, Kristina},
  journal={arXiv preprint arXiv:2103.12149},
  year={2021}
}

@article{DBLP:journals/data/MelotteK21,
  author    = {Sara Melotte and
               Mayank Kejriwal},
  title     = {A Geo-Tagged {COVID-19} Twitter Dataset for 10 North American Metropolitan
               Areas over a 255-Day Period},
  journal   = {Data},
  volume    = {6},
  number    = {6},
  pages     = {64},
  year      = {2021},
  url       = {https://doi.org/10.3390/data6060064},
  doi       = {10.3390/data6060064},
  timestamp = {Thu, 14 Oct 2021 01:00:00 +0200},
  biburl    = {https://dblp.org/rec/journals/data/MelotteK21.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

@InProceedings{10.1007/978-3-030-88361-4_18,
author="Vu, Binh and Knoblock, Craig A. and Szekely, Pedro and Pham, Minh and Pujara, Jay",
editor="Hotho, Andreas and Blomqvist, Eva and Dietze, Stefan and Fokoue, Achille and Ding, Ying and Barnaghi, Payam and Haller, Armin and Dragoni, Mauro and Alani, Harith",
title="A Graph-Based Approach for Inferring Semantic Descriptions of Wikipedia Tables",
booktitle="The Semantic Web -- ISWC 2021",
year="2021",
publisher="Springer International Publishing",
address="Cham",
pages="304--320",
abstract="There are millions of high-quality tables available in Wikipedia. These tables cover many domains and contain useful information. To make use of these tables for data discovery or data integration, we need precise descriptions of the concepts and relationships in the data, known as semantic descriptions. However, creating semantic descriptions is a complex process requiring considerable manual effort and can be error prone. In this paper, we present a novel probabilistic approach for automatically building semantic descriptions of Wikipedia tables. Our approach leverages hyperlinks in a Wikipedia table and existing knowledge in Wikidata to construct a graph of possible relationships in the table and its context, and then it uses collective inference to distinguish genuine and spurious relationships to form the final semantic description. In contrast to existing methods, our solution can handle tables that require complex semantic descriptions of n-ary relations (e.g., the population of a country in a particular year) or implicit contextual values to describe the data accurately. In our empirical evaluation, our approach outperforms state-of-the-art systems on the SemTab2020 dataset and outperforms those systems by as much as 28{\%} in F1 score on a large set of Wikipedia tables.",
isbn="978-3-030-88361-4",
urlLink="https://link.springer.com/chapter/10.1007/978-3-030-88361-4_18",
urlPaper="http://usc-isi-i2.github.io/papers/vu21-springer.pdf",
URLslides = "http://usc-isi-i2.github.io/slides/vu-iswc21-slides.pdf"
}

There are millions of high-quality tables available in Wikipedia. These tables cover many domains and contain useful information. To make use of these tables for data discovery or data integration, we need precise descriptions of the concepts and relationships in the data, known as semantic descriptions. However, creating semantic descriptions is a complex process requiring considerable manual effort and can be error prone. In this paper, we present a novel probabilistic approach for automatically building semantic descriptions of Wikipedia tables. Our approach leverages hyperlinks in a Wikipedia table and existing knowledge in Wikidata to construct a graph of possible relationships in the table and its context, and then it uses collective inference to distinguish genuine and spurious relationships to form the final semantic description. In contrast to existing methods, our solution can handle tables that require complex semantic descriptions of n-ary relations (e.g., the population of a country in a particular year) or implicit contextual values to describe the data accurately. In our empirical evaluation, our approach outperforms state-of-the-art systems on the SemTab2020 dataset and outperforms those systems by as much as 28% in F1 score on a large set of Wikipedia tables.

@inproceedings{vu:iswc21,
    Author = "Vu, Binh and Knoblock, Craig and Szekely, Pedro and Pujara, Jay and Pham, Minh",
    title = "A Graph-based Approach for Inferring Semantic Descriptions of Wikip