Call for Nominations

The Robert J. Glushko Dissertation Prizes in Cognitive Science

Nomination Deadline: January 15, 2018

The Cognitive Science Society and the Glushko-Samuelson Foundation seek nominations for up to five outstanding dissertation prizes in cognitive science. The goals of these prizes are to increase the prominence of cognitive science, and encourage students to engage in interdisciplinary efforts to understand minds and intelligent systems. The hope is that the prizes will recognize and honor young researchers conducting ground-breaking research in cognitive science. The eventual goal is to aid in efforts to bridge between the areas of cognitive science and create theories of general interest to the multiple fields concerned with scientifically understanding the nature of minds and intelligent systems. Promoting a unified cognitive science is consistent with the belief that understanding how minds work will require the synthesis of many different empirical methods, formal tools, and analytic theories. The prize was first begun in 2011, and 2018 will occasion the induction of the eighth group of prize winners. Up-to-date information on the prizes can be found here.

See the full announcement here.

Kristine Yu (UMass Linguistics) will present a talk title “Recovering sentence structure from spoken language” in the Cognitive Bag Lunch series, Wednesday at noon in Tobin 521B. All are welcome!

Jeremy Wolfe  has written a post on the NIH definition of clinical trials on the Psychonomic Society’s website. See the post here: https://featuredcontent.psychonomic.org/the-nih-definition-of-clinical-trials-an-update-for-the-holiday-season/

The UMass Computational Social Science Institute invites you to an engaging (and timely) talk from a visiting postdoc, collaborating with two CSSI affiliates:

Rishab Nithyanand
Ford-Mozilla Open Web Fellow, Data & Society, New York
Friday, December 8, 2017 • 12:30 p.m.-2:00 p.m. (lunch served at 12:15)
Computer Science Building, Room 150/151
Collaboration with CSSI Affiliates Brian Schaffner (Political Science) and Philippa Gill (Computer Science)

Title: Online Political Discourse in the Trump Era

Abstract: We identify general trends in the (in)civility and complexity of political discussions occurring on Reddit between January 2007 and May 2017 – a period spanning both terms of Barack Obama’s presidency and the first 100 days of Donald Trump’s presidency. We then investigate four factors that are frequently hypothesized as having contributed to the declining quality of American political discourse – (1) the rising popularity of Donald Trump, (2) increasing polarization and negative partisanship, (3) the democratization of news media and the rise of fake news, and (4) merging of fringe groups into mainstream political discussions.

Bio: Rishab is currently a Ford-Mozilla Open Web Fellow at Data & Society in New York. His research interests are in the areas of security, privacy, and Internet measurement. More specifically, he likes research geared towards defending anonymity, understanding Internet censorship, measuring aspects of the advertising and tracking ecosystem, and generally understanding the impact of the Internet on the sociopolitical realities of today.

Valentine Hacquard of the University of Maryland will be presenting “Learning what ‘must’ and ‘can’ must and can mean”, Friday December 8th at 3:30 in ILC N400. All are welcome – a reception will follow.

Abstract. The way languages across the world express modality shows both variation and convergence. In some languages, like English, the same modal words (e.g., must) can express different flavors of modality: “Jo must eat fish”, for instance, can express a likelihood that Jo is a fish eater (‘epistemic’ necessity) or an obligation Jo has to eat fish (‘deontic’ necessity). In other languages, modals are strictly monosemous. How do children figure out the modals of their language? What expectations, if any, do they bring to this learning problem? This talk focuses on English-learning children, and asks how they figure out that their modals can be used to express different flavors, what in their linguistic experience might give away modal polysemy, and what linguistic biases might guide this acquisition process.

Chris White will present in the Cognitive Bag Lunch series, Wednesday at noon in Tobin 521B. All are welcome!

Rajesh Ranganath (NYU) will present “Black Box Variational Inference: Scalable, Generic Bayesian Computation and its Applications” in the Machine Learning and Friends Lunch Thursday Nov. 13 at 11:45 am in CS 150. Abstract and bio follow.

Abstract:

Probabilistic generative models are robust to noise, uncover unseen patterns, and make predictions about the future. Probabilistic generative models posit hidden structure to describe data. They have addressed problems in neuroscience, astrophysics, genetics, and medicine. The main computational challenge is computing the hidden structure given the data — posterior inference. For most models of interest, computing the posterior distribution requires approximations like variational inference. Classically, variational inference was feasible to deploy in only a small fraction of models. We develop black box variational inference. Black box variational inference is a variational inference algorithm that is easy to deploy on a broad class of models and has already found use in neuroscience and healthcare. The ideas around black box variational inference also facilitate new kinds of variational methods such as hierarchical variational models. Hierarchical variational models improve the approximation quality of variational inference by building higher-fidelity approximations from coarser ones. Black box variational inference opens the doors to new models and better posterior approximations. Lastly, I will discuss some of the challenges that variational methods face moving forward.

Bio:

Rajesh Ranganath is a postdoc at Columbia University’s Department of Statistics and a research affiliate at MIT’s Institute for Medical Engineering and Science. He will be an assistant professor at the Courant Institute of Mathematical Sciences at NYU starting January 2018. His research interests include approximate inference, model checking, Bayesian nonparametrics, and machine learning for healthcare. Rajesh recently completed his PhD at Princeton with David Blei. Before starting his PhD, Rajesh worked as a software engineer for AMA Capital Management. He obtained his BS and MS from Stanford University with Andrew Ng and Dan Jurafsky. Rajesh has won several awards and fellowships including the NDSEG graduate fellowship and the Porter Ogden Jacobus Fellowship, given to the top four doctoral students at Princeton University.

Screening of AlphaGo
November 30
S131 Integrated Learning Center

Showing at 5pm.
Note this event is not in the Computer Science Building and seating is limited.

With more board configurations than there are atoms in the universe, the ancient Chinese game of Go has long been considered a grand challenge for artificial intelligence. On March 9, 2016, the worlds of Go and artificial intelligence collided in South Korea for an extraordinary best-of-five-game competition, coined The DeepMind Challenge Match. Hundreds of millions of people around the world watched as a legendary Go master took on an unproven AI challenger for the first time in history.

Directed by Greg Kohs with an original score by Academy Award nominee, Hauschka, AlphaGo chronicles a journey from the halls of Oxford, through the backstreets of Bordeaux, past the coding terminals of Google DeepMind in London, and ultimately, to the seven-day tournament in Seoul. As the drama unfolds, more questions emerge: What can artificial intelligence reveal about a 3000-year-old game? What can it teach us about humanity?

https://www.alphagomovie.com

Steven Gilbers of the University of Groningen will be giving a special talk on: “Regional variation in African American English and hip-hop. Why 2Pac’s accent changed over time and why Snoop Dogg and Jay Z have different rap flows”. It will be held Friday Dec. 1 at 10 am in ILC N451. All are welcome! An abstract and bio are below.

Abstract. Relatively little is yet known about how African American English (AAE) regiolects differ from each other. However, we do know regional variation in AAE is salient to many of its speakers, especially those involved with hip-hop culture, in which great importance is assigned to regional identity, and regional accents are a key means of expressing regional identity and affiliation (Morgan, 2001). In hip-hop music, regional variation can also be observed, with different regions’ rap performances being characterized by distinct “flows” (i.e. rhythmic and melodic delivery), possibly due to certain language varieties being better suited for certain flows (Kautny, 2015).

The observations above inform Steven Gilbers’s dissertational research on hip-hop linguistics. During his upcoming talk at UMass Amherst, he will discuss how East Coast and West Coast AAE differ from each other in terms of vowel duration and prosody as well as how these differences are reflected in the rap styles associated with both regions. Moreover, he will discuss how Tupac “2Pac” Shakur – a native New Yorker – acquired a West Coast AAE accent, and how his second dialect acquisition trajectory was influenced by his role in the East Coast/West Coast hip-hop war of the 1990s.

Bio. Steven Gilbers (26) is a hip-hop linguist from Groningen, the Netherlands. His research interests include African American English, hip-hop music, and the sociolinguistics of hip-hop culture. Steven is in the process of writing his doctoral dissertation on second African American English dialect acquisition in relation to regional hip-hop identity at the University of Groningen. Supported by a Fulbright grant, he is currently visiting the United States to conduct an African American English accent perception experiment in New York City and Los Angeles. Aside from his academic endeavors, Steven is also a hip-hop musician, spoken word artist, and co-host of the Kick Knowledge podcast.

Jeff Dubin (ethics) will present in the Cognitive Bag Lunch series, Wednesday at noon in Tobin 521B. All are welcome!