83 |
84 | **Summary:** During the last ten years, network science has emerged as a strongly
85 | interdisciplinary new field of studies in academia. Here, I will trace
86 | the epistemic development of the field examining the mathematical
87 | origins of certain important concepts of network analysis as they were
88 | exported mainly from the rudiments of graph theory. I will discuss a
89 | few examples of schools of mathematics (mostly in USA) starting to
90 | offer courses and to host research on network science. Moreover, I
91 | will present the Wikipedia network stemming from the 3 Wikipedia pages
92 | (Graph theory, Network science, Complex network) and I will discuss
93 | the structure of this Wikipedia network based on the graph-theoretic
94 | notion of minimum dominating sets. The exemplary network that will be
95 | presented here is that of the (direct and) indirect collaborations
96 | among members of the NYUAD Faculty of Mathematics generated by data
97 | extracted from the MathSciNet database of the American Mathematical
98 | Society. Furthermore, I will analyze the bipartite graph of faculty
99 | members and research interests using the tools of Formal Concept
100 | Analysis. Finally, I will model the effect of a network influence
101 | process on the Math collaboration network when research areas are
102 | sources of boundary influence simulations.
103 |
104 |
105 |
106 |
107 |
108 |
109 |
Thursday, `r date`
110 |
11:00am-12:20pm
111 |
Cummings Life Science Center, Room 101
112 |
113 |
114 |
115 |
116 |
117 |
118 |
119 |
120 |
121 | **`r speaker`** is in the Faculty of Northwestern University School of Professional Studies Data Science Program. Currently he is Visiting Professor of Mathematics at the New York University Abu Dhabi (NYUAD), where he is affiliated to the Research Group in Network Science [RGNS](https://sites.google.com/nyu.edu/rgns/home).
122 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: During the last ten years, network science has emerged as a strongly interdisciplinary new field of studies in academia. Here, I will trace the epistemic development of the field examining the mathematical origins of certain important concepts of network analysis as they were exported mainly from the rudiments of graph theory. I will discuss a few examples of schools of mathematics (mostly in USA) starting to offer courses and to host research on network science. Moreover, I will present the Wikipedia network stemming from the 3 Wikipedia pages (Graph theory, Network science, Complex network) and I will discuss the structure of this Wikipedia network based on the graph-theoretic notion of minimum dominating sets. The exemplary network that will be presented here is that of the (direct and) indirect collaborations among members of the NYUAD Faculty of Mathematics generated by data extracted from the MathSciNet database of the American Mathematical Society. Furthermore, I will analyze the bipartite graph of faculty members and research interests using the tools of Formal Concept Analysis. Finally, I will model the effect of a network influence process on the Math collaboration network when research areas are sources of boundary influence simulations.
134 |
135 |
136 |
137 | Thursday, 01/09
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 |
148 |
149 | Moses Boudourides is in the Faculty of Northwestern University School of Professional Studies Data Science Program. Currently he is Visiting Professor of Mathematics at the New York University Abu Dhabi (NYUAD), where he is affiliated to the Research Group in Network Science RGNS.
150 |
161 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
162 |
163 |
164 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
165 |
43 |
44 | **Summary:** During the last ten years, network science has emerged as a strongly
45 | interdisciplinary new field of studies in academia. Here, I will trace
46 | the epistemic development of the field examining the mathematical
47 | origins of certain important concepts of network analysis as they were
48 | exported mainly from the rudiments of graph theory. I will discuss a
49 | few examples of schools of mathematics (mostly in USA) starting to
50 | offer courses and to host research on network science. Moreover, I
51 | will present the Wikipedia network stemming from the 3 Wikipedia pages
52 | (Graph theory, Network science, Complex network) and I will discuss
53 | the structure of this Wikipedia network based on the graph-theoretic
54 | notion of minimum dominating sets. The exemplary network that will be
55 | presented here is that of the (direct and) indirect collaborations
56 | among members of the NYUAD Faculty of Mathematics generated by data
57 | extracted from the MathSciNet database of the American Mathematical
58 | Society. Furthermore, I will analyze the bipartite graph of faculty
59 | members and research interests using the tools of Formal Concept
60 | Analysis. Finally, I will model the effect of a network influence
61 | process on the Math collaboration network when research areas are
62 | sources of boundary influence simulations.
63 |
64 |
65 |
66 |
67 |
68 |
69 |
Thursday, 01/09
70 |
11:00am-12:20pm
71 |
Cummings Life Science Center, Room 101
72 |
73 |
74 |
75 |
76 |
77 |
78 |
79 |
80 |
81 | **Moses Boudourides** is in the Faculty of Northwestern University School of Professional Studies Data Science Program. Currently he is Visiting Professor of Mathematics at the New York University Abu Dhabi (NYUAD), where he is affiliated to the Research Group in Network Science [RGNS](https://sites.google.com/nyu.edu/rgns/home).
82 |
83 |
84 |
85 |
86 |
87 | This week's suggested readings:
88 |
89 | - [Boudourides and Lenis. "Boundary Stimulation of Social Influence Networks."](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-09_Boudourides/Boudourides_Lenis_BoundaryStimulation.pdf)
90 | - [Boudourides and Lenis (2016). "Dominating Sets and Ego-Centered Decompositions in Social Networks." *EPJST* 225:1293-1310.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-09_Boudourides/Boudourides_Lenis_EPJST.pdf)
91 |
92 |
93 |
94 |
95 |
96 | ---
97 |
98 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
79 |
80 | **Summary:** Trust is a fundamental prerequisite in the growth and sustainability
81 | of sharing economy platforms. Many of such platforms rely on
82 | actions that require trust to take place, such as entering a stranger’s
83 | car or sleeping at a stranger’s place. For this reason, understanding,
84 | measuring, and tracking trust can be of great benefit to such platforms,
85 | enabling them to identify trust behaviors, both online and
86 | offline, and identify groups which may benefit from trust-building
87 | interventions. In this work, we present the design and evaluation
88 | of a behavioral framework to measure a user’s propensity to trust
89 | others on Airbnb. We conducted an online experiment with 4,499
90 | Airbnb users in the form of an investment game in order to capture
91 | users’ propensity to trust other users on Airbnb. Then, we used the
92 | experimental data to generate both explanatory and predictive models
93 | of trust propensity. Our contribution is a framework that can
94 | be used to measure trust propensity in sharing economy platforms
95 | via online and offline signals. We discuss which affordances need
96 | to be in place so that sharing economy platforms can get signals of
97 | trust, in addition to how such a framework can be used to inform
98 | design around trust in the short and long term.
99 |
100 |
101 |
102 |
103 |
104 |
105 |
Thursday, `r date`
106 |
11:00am-12:20pm
107 |
Cummings Life Science Center, Room 101
108 |
109 |
110 |
111 |
112 |
113 |
114 |
115 |
116 |
117 | **`r speaker`** is the Lead Trust Scientist at Airbnb and the Associate Director for Computational Social Science at IRiSS, Stanford University. Paolo uses computational techniques and methods to issues of exchange and trust in the “sharing economy”. His published work has appeared on PNAS, The American Journal of Sociology, PLoS 1, and several other academic journals.
118 |
119 |
120 | **Natã M. Barbosa** is a PhD candidate at the University of Illinois at Urbana-Champaign's School of Information Sciences working on data-driven systems for privacy, security, and trust.
121 |
122 |
123 |
124 |
125 | This week's suggested readings:
126 |
127 | - `r readings0`
128 | - **Note**: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
129 |
130 |
131 |
132 |
133 |
134 |
135 | ---
136 |
137 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: Trust is a fundamental prerequisite in the growth and sustainability of sharing economy platforms. Many of such platforms rely on actions that require trust to take place, such as entering a stranger’s car or sleeping at a stranger’s place. For this reason, understanding, measuring, and tracking trust can be of great benefit to such platforms, enabling them to identify trust behaviors, both online and offline, and identify groups which may benefit from trust-building interventions. In this work, we present the design and evaluation of a behavioral framework to measure a user’s propensity to trust others on Airbnb. We conducted an online experiment with 4,499 Airbnb users in the form of an investment game in order to capture users’ propensity to trust other users on Airbnb. Then, we used the experimental data to generate both explanatory and predictive models of trust propensity. Our contribution is a framework that can be used to measure trust propensity in sharing economy platforms via online and offline signals. We discuss which affordances need to be in place so that sharing economy platforms can get signals of trust, in addition to how such a framework can be used to inform design around trust in the short and long term.
134 |
135 |
136 |
137 | Thursday, 01/16
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 |
148 |
149 | Paolo Parigi is the Lead Trust Scientist at Airbnb and the Associate Director for Computational Social Science at IRiSS, Stanford University. Paolo uses computational techniques and methods to issues of exchange and trust in the “sharing economy”. His published work has appeared on PNAS, The American Journal of Sociology, PLoS 1, and several other academic journals.
150 |
151 |
Natã M. Barbosa is a PhD candidate at the University of Illinois at Urbana-Champaign’s School of Information Sciences working on data-driven systems for privacy, security, and trust.
Note: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
157 |
158 |
159 |
160 |
161 |
162 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
163 |
164 |
165 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
166 |
43 |
44 | **Summary:** Trust is a fundamental prerequisite in the growth and sustainability
45 | of sharing economy platforms. Many of such platforms rely on
46 | actions that require trust to take place, such as entering a stranger’s
47 | car or sleeping at a stranger’s place. For this reason, understanding,
48 | measuring, and tracking trust can be of great benefit to such platforms,
49 | enabling them to identify trust behaviors, both online and
50 | offline, and identify groups which may benefit from trust-building
51 | interventions. In this work, we present the design and evaluation
52 | of a behavioral framework to measure a user’s propensity to trust
53 | others on Airbnb. We conducted an online experiment with 4,499
54 | Airbnb users in the form of an investment game in order to capture
55 | users’ propensity to trust other users on Airbnb. Then, we used the
56 | experimental data to generate both explanatory and predictive models
57 | of trust propensity. Our contribution is a framework that can
58 | be used to measure trust propensity in sharing economy platforms
59 | via online and offline signals. We discuss which affordances need
60 | to be in place so that sharing economy platforms can get signals of
61 | trust, in addition to how such a framework can be used to inform
62 | design around trust in the short and long term.
63 |
64 |
65 |
66 |
67 |
68 |
69 |
Thursday, 01/16
70 |
11:00am-12:20pm
71 |
Cummings Life Science Center, Room 101
72 |
73 |
74 |
75 |
76 |
77 |
78 |
79 |
80 |
81 | **Paolo Parigi** is the Lead Trust Scientist at Airbnb and the Associate Director for Computational Social Science at IRiSS, Stanford University. Paolo uses computational techniques and methods to issues of exchange and trust in the “sharing economy”. His published work has appeared on PNAS, The American Journal of Sociology, PLoS 1, and several other academic journals.
82 |
83 |
84 | **Natã M. Barbosa** is a PhD candidate at the University of Illinois at Urbana-Champaign's School of Information Sciences working on data-driven systems for privacy, security, and trust.
85 |
86 |
87 |
88 |
89 | This week's suggested readings:
90 |
91 | - [Nata Barbosa, Emily Sun, Judd Antin, and Paolo Parigi 2018. 'Designing for Trust: A Behavioral Framework for Sharing Economy Platforms'](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-16_Parigi/)
92 | - **Note**: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
93 |
94 |
95 |
96 |
97 |
98 |
99 | ---
100 |
101 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
83 |
84 | **Summary:** Setbacks are an integral part of a scientific career, yet little is known about their long-term effects. Here we examine junior scientists applying for National Institutes of Health R01 grants. By focusing on proposals that fell just below and just above the funding threshold, we compare near-miss with narrow-win applicants, and find that an early-career setback has powerful, opposing effects. On the one hand, it significantly increases attrition, predicting more than a 10% chance of disappearing permanently from the NIH system. Yet, despite an early setback, individuals with near misses systematically outperform those with narrow wins in the longer run. Moreover, this performance advantage seems to go beyond a screening mechanism, suggesting early-career setback appears to cause a performance improvement among those who persevere. Overall, these findings are consistent with the concept that “what doesn’t kill me makes me stronger,” which may have broad implications for identifying, training and nurturing junior scientists. I also discuss a second project in which we model how human achievements are often preceded by repeated failures to illuminate the mechanisms that govern the dynamics of failure. Building on previous research relating to innovation, human dynamics and learning, we develop a simple one-parameter model that mimics how successful future attempts build on past efforts. Solving this model analytically suggests that a phase transition separates the dynamics of failure into regions of progression or stagnation and predicts that, near the critical threshold, agents who share similar characteristics and learning strategies may experience fundamentally different outcomes following failures. Above the critical point, agents exploit incremental refinements to systematically advance towards success, whereas below it, they explore disjoint opportunities without a pattern of improvement. The model makes several empirically testable predictions, demonstrating that those who eventually succeed and those who do not may initially appear similar, but can be characterized by fundamentally distinct failure dynamics in terms of the efficiency and quality associated with each subsequent attempt. We validated these with large-scale data from three disparate domains—scientists attempting to obtain National Institutes of Health (NIH) grants to fund their research, entrepreneurs seeking to successfully exit their startup ventures, and terrorist organizations trying claim casualties in violent attacks. Our findings unveil detectable yet previously unknown early signals that enable us to identify failure dynamics that predict ultimate success or failure. Given the ubiquitous nature of failure and the paucity of quantitative approaches to understand it, these results represent an initial step towards deeper understanding of the complex dynamics underlying failure.
85 |
86 |
87 |
88 |
89 |
90 |
91 |
Thursday, `r date`
92 |
11:00am-12:20pm
93 |
Cummings Life Science Center, Room 101
94 |
95 |
96 |
97 |
A light lunch will be provided by `r vendor`.
98 |
99 |
100 |
101 |
102 |
103 | **`r speaker`** is an Associate Professor of Management and Organizations at the Kellogg School of Management, and (by courtesy) the McCormick School of Engineering, at Northwestern University. At Kellogg, he is the Founding Director of the Center for Science of Science and Innovation (CSSI). He is also a core faculty at the Northwestern Institute on Complex Systems (NICO). His current research focus is on Science of Science, a quest to turn the scientific methods and curiosities upon ourselves, hoping to use and develop tools from complexity sciences and artificial intelligence to broadly explore the opportunities for innovation and promises of prosperity offered by the recent data explosion in science. Dashun is a recipient of the AFOSR Young Investigator award (2016) and Poets & Quants Best 40 Under 40 Professors (2019).
104 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: Setbacks are an integral part of a scientific career, yet little is known about their long-term effects. Here we examine junior scientists applying for National Institutes of Health R01 grants. By focusing on proposals that fell just below and just above the funding threshold, we compare near-miss with narrow-win applicants, and find that an early-career setback has powerful, opposing effects. On the one hand, it significantly increases attrition, predicting more than a 10% chance of disappearing permanently from the NIH system. Yet, despite an early setback, individuals with near misses systematically outperform those with narrow wins in the longer run. Moreover, this performance advantage seems to go beyond a screening mechanism, suggesting early-career setback appears to cause a performance improvement among those who persevere. Overall, these findings are consistent with the concept that “what doesn’t kill me makes me stronger,” which may have broad implications for identifying, training and nurturing junior scientists. I also discuss a second project in which we model how human achievements are often preceded by repeated failures to illuminate the mechanisms that govern the dynamics of failure. Building on previous research relating to innovation, human dynamics and learning, we develop a simple one-parameter model that mimics how successful future attempts build on past efforts. Solving this model analytically suggests that a phase transition separates the dynamics of failure into regions of progression or stagnation and predicts that, near the critical threshold, agents who share similar characteristics and learning strategies may experience fundamentally different outcomes following failures. Above the critical point, agents exploit incremental refinements to systematically advance towards success, whereas below it, they explore disjoint opportunities without a pattern of improvement. The model makes several empirically testable predictions, demonstrating that those who eventually succeed and those who do not may initially appear similar, but can be characterized by fundamentally distinct failure dynamics in terms of the efficiency and quality associated with each subsequent attempt. We validated these with large-scale data from three disparate domains—scientists attempting to obtain National Institutes of Health (NIH) grants to fund their research, entrepreneurs seeking to successfully exit their startup ventures, and terrorist organizations trying claim casualties in violent attacks. Our findings unveil detectable yet previously unknown early signals that enable us to identify failure dynamics that predict ultimate success or failure. Given the ubiquitous nature of failure and the paucity of quantitative approaches to understand it, these results represent an initial step towards deeper understanding of the complex dynamics underlying failure.
134 |
135 |
136 |
137 | Thursday, 01/23
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 | A light lunch will be provided by Papa Johns.
148 |
149 |
150 |
151 | Dashun Wang is an Associate Professor of Management and Organizations at the Kellogg School of Management, and (by courtesy) the McCormick School of Engineering, at Northwestern University. At Kellogg, he is the Founding Director of the Center for Science of Science and Innovation (CSSI). He is also a core faculty at the Northwestern Institute on Complex Systems (NICO). His current research focus is on Science of Science, a quest to turn the scientific methods and curiosities upon ourselves, hoping to use and develop tools from complexity sciences and artificial intelligence to broadly explore the opportunities for innovation and promises of prosperity offered by the recent data explosion in science. Dashun is a recipient of the AFOSR Young Investigator award (2016) and Poets & Quants Best 40 Under 40 Professors (2019).
152 |
163 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
164 |
165 |
166 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
167 |
43 |
44 | **Summary:** Setbacks are an integral part of a scientific career, yet little is known about their long-term effects. Here we examine junior scientists applying for National Institutes of Health R01 grants. By focusing on proposals that fell just below and just above the funding threshold, we compare near-miss with narrow-win applicants, and find that an early-career setback has powerful, opposing effects. On the one hand, it significantly increases attrition, predicting more than a 10% chance of disappearing permanently from the NIH system. Yet, despite an early setback, individuals with near misses systematically outperform those with narrow wins in the longer run. Moreover, this performance advantage seems to go beyond a screening mechanism, suggesting early-career setback appears to cause a performance improvement among those who persevere. Overall, these findings are consistent with the concept that “what doesn’t kill me makes me stronger,” which may have broad implications for identifying, training and nurturing junior scientists. I also discuss a second project in which we model how human achievements are often preceded by repeated failures to illuminate the mechanisms that govern the dynamics of failure. Building on previous research relating to innovation, human dynamics and learning, we develop a simple one-parameter model that mimics how successful future attempts build on past efforts. Solving this model analytically suggests that a phase transition separates the dynamics of failure into regions of progression or stagnation and predicts that, near the critical threshold, agents who share similar characteristics and learning strategies may experience fundamentally different outcomes following failures. Above the critical point, agents exploit incremental refinements to systematically advance towards success, whereas below it, they explore disjoint opportunities without a pattern of improvement. The model makes several empirically testable predictions, demonstrating that those who eventually succeed and those who do not may initially appear similar, but can be characterized by fundamentally distinct failure dynamics in terms of the efficiency and quality associated with each subsequent attempt. We validated these with large-scale data from three disparate domains—scientists attempting to obtain National Institutes of Health (NIH) grants to fund their research, entrepreneurs seeking to successfully exit their startup ventures, and terrorist organizations trying claim casualties in violent attacks. Our findings unveil detectable yet previously unknown early signals that enable us to identify failure dynamics that predict ultimate success or failure. Given the ubiquitous nature of failure and the paucity of quantitative approaches to understand it, these results represent an initial step towards deeper understanding of the complex dynamics underlying failure.
45 |
46 |
47 |
48 |
49 |
50 |
51 |
Thursday, 01/23
52 |
11:00am-12:20pm
53 |
Cummings Life Science Center, Room 101
54 |
55 |
56 |
57 |
A light lunch will be provided by Papa Johns.
58 |
59 |
60 |
61 |
62 |
63 | **Dashun Wang** is an Associate Professor of Management and Organizations at the Kellogg School of Management, and (by courtesy) the McCormick School of Engineering, at Northwestern University. At Kellogg, he is the Founding Director of the Center for Science of Science and Innovation (CSSI). He is also a core faculty at the Northwestern Institute on Complex Systems (NICO). His current research focus is on Science of Science, a quest to turn the scientific methods and curiosities upon ourselves, hoping to use and develop tools from complexity sciences and artificial intelligence to broadly explore the opportunities for innovation and promises of prosperity offered by the recent data explosion in science. Dashun is a recipient of the AFOSR Young Investigator award (2016) and Poets & Quants Best 40 Under 40 Professors (2019).
64 |
65 |
66 |
67 |
68 |
69 | This week's suggested readings:
70 |
71 | - [Wang, Yang, Benjamin F. Jones, and Dashun Wang. 2019. "Early-career setback and future career impact." *Nature Communications* 10:4331. DOI:10.1038/s41467-019-12189-3.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-23_Wang/s41467-019-12189-3.pdf)
72 | - [Yin, Yian, Yang Wang, James A. Evans, and Dashun Wang. 2019. "Quantifying the dynamics of failure across science, startups and security." *Nature* 575:190–194.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-23_Wang/s41586-019-1725-y.pdf)
73 |
74 |
75 |
76 |
77 |
78 | ---
79 |
80 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
79 |
80 | **Summary:** Research on belief formation has produced contradictory findings on whether and when
81 | communication between group members will improve the accuracy of estimations such as
82 | economic forecasts, medical diagnoses, and job candidate assessments. While some evidence
83 | suggests that carefully mediated processes (i.e., the “Delphi method”) produce more accurate
84 | beliefs than unstructured discussion, others argue that unstructured discussion outperforms
85 | mediated processes. Still others argue that independent individuals produce the most accurate
86 | beliefs. This paper shows how network theories of belief formation can resolve these
87 | inconsistencies. Emergent network structures of influence—even in groups with no apparent
88 | structure, such as committees—interact with the pre-discussion belief distribution to moderate the
89 | effect of communication on belief formation. As a result, communication sometimes increases and
90 | sometimes decreases the accuracy of the average belief in a group. The effects differ for mediated
91 | processes and unstructured communication, such that the relative benefit of each communication
92 | format depends on both group dynamics as well as the statistical properties of pre-interaction
93 | beliefs. These results resolve contradictions in previous research and offer practical
94 | recommendations for teams and organizations.
95 |
96 |
97 |
98 |
99 |
100 |
101 |
Thursday, `r date`
102 |
11:00am-12:20pm
103 |
Cummings Life Science Center, Room 101
104 |
105 |
106 |
107 |
108 |
109 |
110 |
111 |
112 |
113 | **`r speaker`** is a postdoctoral fellow at the Kellogg School of Management, Northwestern University, and a researcher in residence at the Northwestern Institute on Complex Systems. Joshua completed his PhD at the Annenberg School for Communication, University of Pennsylvania. Prior to graduate school, he worked professionally in mediation and conflict resolution and now serves as a pro-bono mediator with the Chicago Conflict Resolution Center. His research has been published in Science, Proceedings of the National Academy of Sciences, and Harvard Business Review.
114 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: Research on belief formation has produced contradictory findings on whether and when communication between group members will improve the accuracy of estimations such as economic forecasts, medical diagnoses, and job candidate assessments. While some evidence suggests that carefully mediated processes (i.e., the “Delphi method”) produce more accurate beliefs than unstructured discussion, others argue that unstructured discussion outperforms mediated processes. Still others argue that independent individuals produce the most accurate beliefs. This paper shows how network theories of belief formation can resolve these inconsistencies. Emergent network structures of influence—even in groups with no apparent structure, such as committees—interact with the pre-discussion belief distribution to moderate the effect of communication on belief formation. As a result, communication sometimes increases and sometimes decreases the accuracy of the average belief in a group. The effects differ for mediated processes and unstructured communication, such that the relative benefit of each communication format depends on both group dynamics as well as the statistical properties of pre-interaction beliefs. These results resolve contradictions in previous research and offer practical recommendations for teams and organizations.
134 |
135 |
136 |
137 | Thursday, 01/30
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 |
148 |
149 | Joshua Becker is a postdoctoral fellow at the Kellogg School of Management, Northwestern University, and a researcher in residence at the Northwestern Institute on Complex Systems. Joshua completed his PhD at the Annenberg School for Communication, University of Pennsylvania. Prior to graduate school, he worked professionally in mediation and conflict resolution and now serves as a pro-bono mediator with the Chicago Conflict Resolution Center. His research has been published in Science, Proceedings of the National Academy of Sciences, and Harvard Business Review.
150 |
160 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
161 |
162 |
163 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
164 |
43 |
44 | **Summary:** Research on belief formation has produced contradictory findings on whether and when
45 | communication between group members will improve the accuracy of estimations such as
46 | economic forecasts, medical diagnoses, and job candidate assessments. While some evidence
47 | suggests that carefully mediated processes (i.e., the “Delphi method”) produce more accurate
48 | beliefs than unstructured discussion, others argue that unstructured discussion outperforms
49 | mediated processes. Still others argue that independent individuals produce the most accurate
50 | beliefs. This paper shows how network theories of belief formation can resolve these
51 | inconsistencies. Emergent network structures of influence—even in groups with no apparent
52 | structure, such as committees—interact with the pre-discussion belief distribution to moderate the
53 | effect of communication on belief formation. As a result, communication sometimes increases and
54 | sometimes decreases the accuracy of the average belief in a group. The effects differ for mediated
55 | processes and unstructured communication, such that the relative benefit of each communication
56 | format depends on both group dynamics as well as the statistical properties of pre-interaction
57 | beliefs. These results resolve contradictions in previous research and offer practical
58 | recommendations for teams and organizations.
59 |
60 |
61 |
62 |
63 |
64 |
65 |
Thursday, 01/30
66 |
11:00am-12:20pm
67 |
Cummings Life Science Center, Room 101
68 |
69 |
70 |
71 |
72 |
73 |
74 |
75 |
76 |
77 | **Joshua Becker** is a postdoctoral fellow at the Kellogg School of Management, Northwestern University, and a researcher in residence at the Northwestern Institute on Complex Systems. Joshua completed his PhD at the Annenberg School for Communication, University of Pennsylvania. Prior to graduate school, he worked professionally in mediation and conflict resolution and now serves as a pro-bono mediator with the Chicago Conflict Resolution Center. His research has been published in Science, Proceedings of the National Academy of Sciences, and Harvard Business Review.
78 |
79 |
80 |
81 |
82 |
83 | This week's suggested readings:
84 |
85 | - [Becker, Joshua. 2019. "Network Structures of Collective Intelligence: The Contingent Benefits of Group Discussion." Working Paper, August 31, 2019.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/01-30_Becker/Becker_Network_Structures_of_Collective_Intelligence.pdf)
86 |
87 |
88 |
89 |
90 |
91 | ---
92 |
93 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
79 |
80 | **Summary:** In recent decades women’s educational attainment has increasingly come to surpass men’s. While recent work made strides in incorporating cultural models of gender into accounts of this reversal, it has implicitly assumed that these models have themselves remained substantively constant. In this article, we apply neural network word embeddings to a 200-million-word corpus of American print media (1930-2009) to examine whether and how these cultural models have changed. Our approach lets us estimate the extent to which each of over 10,000 English words occurs in feminine versus masculine contexts at different time points. We use this to track the changing gendered connotations of terms related to school, studying, socio-behavioral skills, behavioral problems, intelligence, and unintelligence. Our results point to three ideal-typical trajectories: (1) socio-behavioral skills and behavioral problems start out significantly feminine and masculine, respectively, and retain those connotations throughout the entire time period; (2) school and studying start out without significant gendered connotations, but finish with significant feminine connotations; (3) intelligence and unintelligence start out neutral or slightly feminine but finish significantly masculine. There is an exceptionally strong negative correlation between trends (2) and (3), which leads us to speculate that both changes are part of one overarching cultural shift.
81 |
82 |
83 |
84 |
85 |
86 |
87 |
Thursday, `r date`
88 |
11:00am-12:20pm
89 |
Cummings Life Science Center, Room 101
90 |
91 |
92 |
93 |
94 |
95 |
96 |
97 |
98 |
99 | **`r speaker`**'s research focuses on culture, cognition, methodology, and public opinion. He examines the supra-individual aspects of attitudes, tastes, and cognitive representations, with a special focus on political views. He is broadly interested in the society-wide distribution of these cultural elements, and the social and cognitive processes that give rise to this distribution. He draws on network analysis, statistics, and computer science to develop novel methods for these investigations. In a separate research stream, he studies the effects of political disagreement on social network structure. His work has appeared in the *American Journal of Sociology*, *Sociological Science*, and *Political Psychology*.
100 |
101 |
102 |
103 |
104 |
105 | This week's suggested readings:
106 |
107 | - `r readings0`
108 | - **Note**: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
109 |
110 |
111 |
112 | ---
113 |
114 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: In recent decades women’s educational attainment has increasingly come to surpass men’s. While recent work made strides in incorporating cultural models of gender into accounts of this reversal, it has implicitly assumed that these models have themselves remained substantively constant. In this article, we apply neural network word embeddings to a 200-million-word corpus of American print media (1930-2009) to examine whether and how these cultural models have changed. Our approach lets us estimate the extent to which each of over 10,000 English words occurs in feminine versus masculine contexts at different time points. We use this to track the changing gendered connotations of terms related to school, studying, socio-behavioral skills, behavioral problems, intelligence, and unintelligence. Our results point to three ideal-typical trajectories: (1) socio-behavioral skills and behavioral problems start out significantly feminine and masculine, respectively, and retain those connotations throughout the entire time period; (2) school and studying start out without significant gendered connotations, but finish with significant feminine connotations; (3) intelligence and unintelligence start out neutral or slightly feminine but finish significantly masculine. There is an exceptionally strong negative correlation between trends (2) and (3), which leads us to speculate that both changes are part of one overarching cultural shift.
134 |
135 |
136 |
137 | Thursday, 02/06
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 |
148 |
149 | Andrei Boutyline’s research focuses on culture, cognition, methodology, and public opinion. He examines the supra-individual aspects of attitudes, tastes, and cognitive representations, with a special focus on political views. He is broadly interested in the society-wide distribution of these cultural elements, and the social and cognitive processes that give rise to this distribution. He draws on network analysis, statistics, and computer science to develop novel methods for these investigations. In a separate research stream, he studies the effects of political disagreement on social network structure. His work has appeared in the American Journal of Sociology, Sociological Science, and Political Psychology.
150 |
Note: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
156 |
157 |
158 |
159 |
160 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
161 |
162 |
163 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
164 |
43 |
44 | **Summary:** In recent decades women’s educational attainment has increasingly come to surpass men’s. While recent work made strides in incorporating cultural models of gender into accounts of this reversal, it has implicitly assumed that these models have themselves remained substantively constant. In this article, we apply neural network word embeddings to a 200-million-word corpus of American print media (1930-2009) to examine whether and how these cultural models have changed. Our approach lets us estimate the extent to which each of over 10,000 English words occurs in feminine versus masculine contexts at different time points. We use this to track the changing gendered connotations of terms related to school, studying, socio-behavioral skills, behavioral problems, intelligence, and unintelligence. Our results point to three ideal-typical trajectories: (1) socio-behavioral skills and behavioral problems start out significantly feminine and masculine, respectively, and retain those connotations throughout the entire time period; (2) school and studying start out without significant gendered connotations, but finish with significant feminine connotations; (3) intelligence and unintelligence start out neutral or slightly feminine but finish significantly masculine. There is an exceptionally strong negative correlation between trends (2) and (3), which leads us to speculate that both changes are part of one overarching cultural shift.
45 |
46 |
47 |
48 |
49 |
50 |
51 |
Thursday, 02/06
52 |
11:00am-12:20pm
53 |
Cummings Life Science Center, Room 101
54 |
55 |
56 |
57 |
58 |
59 |
60 |
61 |
62 |
63 | **Andrei Boutyline**'s research focuses on culture, cognition, methodology, and public opinion. He examines the supra-individual aspects of attitudes, tastes, and cognitive representations, with a special focus on political views. He is broadly interested in the society-wide distribution of these cultural elements, and the social and cognitive processes that give rise to this distribution. He draws on network analysis, statistics, and computer science to develop novel methods for these investigations. In a separate research stream, he studies the effects of political disagreement on social network structure. His work has appeared in the *American Journal of Sociology*, *Sociological Science*, and *Political Psychology*.
64 |
65 |
66 |
67 |
68 |
69 | This week's suggested readings:
70 |
71 | - [Boutyline, Andrei, Alina Arseniev-Koehler, and Devin J. Cornell. 2020. "School, Studying, and Smarts: The Gender of Education Across 80 Years of American Print Media, 1930-2009," Working Paper, January 29, 2020.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/02-06_Becker/)
72 | - **Note**: This reading was privately sent to MACSS students and faculty. Please do not share or distribute the file.
73 |
74 |
75 |
76 | ---
77 |
78 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
83 |
84 | **Summary:** In this workshop, I will introduce a hierarchical Dirichlet process of generalized linear models in which the latent heterogeneity in the effect of individual-level covariates depends on context-level features. Such a model is important in political analyses, for instance, when the data comes from different countries and the latent heterogeneity in political behavior can be a function of country-level characteristics. The supporting paper proposes a semi-parametric Bayesian approach, a Gibbs sampler for the general model, a special Gibbs sampler for Gaussian outcome variables, and a Hamiltonian Monte Carlo within Gibbs to handle discrete outcome variables. I demonstrate the importance of accounting for latent heterogeneity with a Monte Carlo exercise and with two applications that replicate recent scholarly work. I demonstrate how the proposed allow us to identify latent structures in public political polarization.
85 |
86 |
87 |
88 |
89 |
90 |
91 |
Thursday, `r date`
92 |
11:00am-12:20pm
93 |
Cummings Life Science Center, Room 101
94 |
95 |
96 |
97 |
98 |
99 |
100 |
101 |
102 |
103 | **`r speaker`** is an Assistant Instructional Professor in the Masters in Computational Social Science program and a Political Scientist with expertise in OECD and Latin America countries. He holds a PhD degree in Political Science and Scientific Computing from the University of Michigan, Ann Arbor, and an MA degree in Statistics from the same university. Dr. Ferrari is interested in a wide range of topics in computational social sciences, comparative politics, and political methodology. He teaches courses on Computational Methods for Political Science, Advanced Machine Learning, Bayesian Statistics, and Introduction to Computer Science. His doctoral research proposes innovative hierarchical unsupervised learning methods to estimate latent interactions in observational and experimental studies and to measure the polarization of policy preferences. Broadly, his substantive research combines political economy, political sociology, and social cognition approaches to study the formation of political preferences. In his recent research, he examines the connections between people's socioeconomic conditions, cognitive perceptions about the socioeconomic environment, and political opinions.
104 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: In this workshop, I will introduce a hierarchical Dirichlet process of generalized linear models in which the latent heterogeneity in the effect of individual-level covariates depends on context-level features. Such a model is important in political analyses, for instance, when the data comes from different countries and the latent heterogeneity in political behavior can be a function of country-level characteristics. The supporting paper proposes a semi-parametric Bayesian approach, a Gibbs sampler for the general model, a special Gibbs sampler for Gaussian outcome variables, and a Hamiltonian Monte Carlo within Gibbs to handle discrete outcome variables. I demonstrate the importance of accounting for latent heterogeneity with a Monte Carlo exercise and with two applications that replicate recent scholarly work. I demonstrate how the proposed allow us to identify latent structures in public political polarization.
134 |
135 |
136 |
137 | Thursday, 02/27
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | Cummings Life Science Center, Room 101
144 |
145 |
146 |
147 |
148 |
149 | Diogo Ferrari is an Assistant Instructional Professor in the Masters in Computational Social Science program and a Political Scientist with expertise in OECD and Latin America countries. He holds a PhD degree in Political Science and Scientific Computing from the University of Michigan, Ann Arbor, and an MA degree in Statistics from the same university. Dr. Ferrari is interested in a wide range of topics in computational social sciences, comparative politics, and political methodology. He teaches courses on Computational Methods for Political Science, Advanced Machine Learning, Bayesian Statistics, and Introduction to Computer Science. His doctoral research proposes innovative hierarchical unsupervised learning methods to estimate latent interactions in observational and experimental studies and to measure the polarization of policy preferences. Broadly, his substantive research combines political economy, political sociology, and social cognition approaches to study the formation of political preferences. In his recent research, he examines the connections between people’s socioeconomic conditions, cognitive perceptions about the socioeconomic environment, and political opinions.
150 |
160 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
161 |
162 |
163 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
164 |
43 |
44 | **Summary:** In this workshop, I will introduce a hierarchical Dirichlet process of generalized linear models in which the latent heterogeneity in the effect of individual-level covariates depends on context-level features. Such a model is important in political analyses, for instance, when the data comes from different countries and the latent heterogeneity in political behavior can be a function of country-level characteristics. The supporting paper proposes a semi-parametric Bayesian approach, a Gibbs sampler for the general model, a special Gibbs sampler for Gaussian outcome variables, and a Hamiltonian Monte Carlo within Gibbs to handle discrete outcome variables. I demonstrate the importance of accounting for latent heterogeneity with a Monte Carlo exercise and with two applications that replicate recent scholarly work. I demonstrate how the proposed allow us to identify latent structures in public political polarization.
45 |
46 |
47 |
48 |
49 |
50 |
51 |
Thursday, 02/27
52 |
11:00am-12:20pm
53 |
Cummings Life Science Center, Room 101
54 |
55 |
56 |
57 |
58 |
59 |
60 |
61 |
62 |
63 | **Diogo Ferrari** is an Assistant Instructional Professor in the Masters in Computational Social Science program and a Political Scientist with expertise in OECD and Latin America countries. He holds a PhD degree in Political Science and Scientific Computing from the University of Michigan, Ann Arbor, and an MA degree in Statistics from the same university. Dr. Ferrari is interested in a wide range of topics in computational social sciences, comparative politics, and political methodology. He teaches courses on Computational Methods for Political Science, Advanced Machine Learning, Bayesian Statistics, and Introduction to Computer Science. His doctoral research proposes innovative hierarchical unsupervised learning methods to estimate latent interactions in observational and experimental studies and to measure the polarization of policy preferences. Broadly, his substantive research combines political economy, political sociology, and social cognition approaches to study the formation of political preferences. In his recent research, he examines the connections between people's socioeconomic conditions, cognitive perceptions about the socioeconomic environment, and political opinions.
64 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
83 |
84 | **Summary:** Innovations have been achieved by combining existing knowledge with novel additions. The combinatorial nature of innovation is mainly governed by exploitation and exploration, behaviors that reuse known combinations and discover new possibilities. The trade-off between them, a well-known dilemma that innovators face, is not only affected by, but also change knowledge space where innovations occur. To investigate how the trade-off contributes to knowledge space, we present a network model that sub-linearly grows by following Heaps' law and creates links with adjacent possibles, and compare model networks to empirical knowledge spaces constructed from patents and research papers. With the balanced trade-off, the model reproduces structural properties found across empirical knowledge spaces, including broad strength distribution, significant local clustering, and especially modular structure that is often considered taken for granted. If either exploitation or exploration dominates the trade-off, these structures could not be obtained. Interestingly, the empirical spaces appear to evolve with the balanced trade-off, suggesting collective behaviors that balance exploitation and exploration naturally shape knowledge modules. Our network model not only explains why modular structure exists within knowledge spaces but also provides insights into the co-evolution of microscopic behaviors and modular structure.
85 |
86 |
87 |
88 |
89 |
90 |
91 |
Friday, `r date`
92 |
11:00am-12:20pm
93 |
1155 E 60TH ST, Room 289A
94 |
95 |
96 |
97 |
A light reception will be provided by `r vendor` following the talk. The reception will be held in the new MACSS lounge, 1155 E 60TH ST, Room 224.
98 |
99 |
100 |
101 |
102 |
103 | **`r speaker`** is an Assistant Professor of Management & Organization Department at the Kellogg School of Management, and a core faculty at NICO, the Northwestern Institute on Complex Systems. She is also Royal Society of Arts fellow, and an external fellow at London Mathematical Laboratory, London, UK. Prior to joining Kellogg, she worked at University of Oxford, Harvard University, and MIT Media Lab, and Santa Fe Institute, as a research fellow. Hyejin received her PhD in Physics in 2011 from Korea Advanced Institute of Science and Technology (KAIST). She was a Principal Investigator of the project a National Science Foundation grant (USA) to study Technological Change from the Map of Capabilities.
104 |
105 | Her research interests are to understand the interplay between technological innovation and socio-economic systems (urbanisation, economic diversity and specialisation, invention activity, future of work). Her highly interdisciplinary approach often results in broad collaborations ranging from mathematicians, computer scientists, economists, sociologists, anthropologists, to archeologists. Her work has been published in general audience such as Nature communication, and PNAS, as well as top specialized journals such Physics Review Letter, and Evolutionary Anthropology, and has been featured in The Econonmist, Forbes, The Guardian, WIRED, Scientific America, MIT Technonlogy Review, among other major global media outlets. Her goal is to develop a theoretical, yet empirically grounded, framework that will enable us to turn the increasing volumes of data into scientific insights and well-designed policies, an approach known as computational social science. The mathematical tools and computational methods that are used include scaling theory, spatial analysis (including percolation theory, information theory and fractal dimension analysis), statistics, and network theory.
106 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
Summary: Innovations have been achieved by combining existing knowledge with novel additions. The combinatorial nature of innovation is mainly governed by exploitation and exploration, behaviors that reuse known combinations and discover new possibilities. The trade-off between them, a well-known dilemma that innovators face, is not only affected by, but also change knowledge space where innovations occur. To investigate how the trade-off contributes to knowledge space, we present a network model that sub-linearly grows by following Heaps’ law and creates links with adjacent possibles, and compare model networks to empirical knowledge spaces constructed from patents and research papers. With the balanced trade-off, the model reproduces structural properties found across empirical knowledge spaces, including broad strength distribution, significant local clustering, and especially modular structure that is often considered taken for granted. If either exploitation or exploration dominates the trade-off, these structures could not be obtained. Interestingly, the empirical spaces appear to evolve with the balanced trade-off, suggesting collective behaviors that balance exploitation and exploration naturally shape knowledge modules. Our network model not only explains why modular structure exists within knowledge spaces but also provides insights into the co-evolution of microscopic behaviors and modular structure.
134 |
135 |
136 |
137 | Friday, 03/06
138 |
139 |
140 | 11:00am-12:20pm
141 |
142 |
143 | 1155 E 60TH ST, Room 289A
144 |
145 |
146 |
147 | A light reception will be provided by Northern Taste following the talk. The reception will be held in the new MACSS lounge, 1155 E 60TH ST, Room 224.
148 |
149 |
150 |
151 |
Hyejin Youn is an Assistant Professor of Management & Organization Department at the Kellogg School of Management, and a core faculty at NICO, the Northwestern Institute on Complex Systems. She is also Royal Society of Arts fellow, and an external fellow at London Mathematical Laboratory, London, UK. Prior to joining Kellogg, she worked at University of Oxford, Harvard University, and MIT Media Lab, and Santa Fe Institute, as a research fellow. Hyejin received her PhD in Physics in 2011 from Korea Advanced Institute of Science and Technology (KAIST). She was a Principal Investigator of the project a National Science Foundation grant (USA) to study Technological Change from the Map of Capabilities.
152 | Her research interests are to understand the interplay between technological innovation and socio-economic systems (urbanisation, economic diversity and specialisation, invention activity, future of work). Her highly interdisciplinary approach often results in broad collaborations ranging from mathematicians, computer scientists, economists, sociologists, anthropologists, to archeologists. Her work has been published in general audience such as Nature communication, and PNAS, as well as top specialized journals such Physics Review Letter, and Evolutionary Anthropology, and has been featured in The Econonmist, Forbes, The Guardian, WIRED, Scientific America, MIT Technonlogy Review, among other major global media outlets. Her goal is to develop a theoretical, yet empirically grounded, framework that will enable us to turn the increasing volumes of data into scientific insights and well-designed policies, an approach known as computational social science. The mathematical tools and computational methods that are used include scaling theory, spatial analysis (including percolation theory, information theory and fractal dimension analysis), statistics, and network theory.
153 |
154 |
163 | The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
164 |
165 |
166 | Students in the Masters of Computational Social Science program are expected to attend and join the discussion by posting a comment on the issues page of the workshop’s public repository on GitHub.
167 |
43 |
44 | **Summary:** Innovations have been achieved by combining existing knowledge with novel additions. The combinatorial nature of innovation is mainly governed by exploitation and exploration, behaviors that reuse known combinations and discover new possibilities. The trade-off between them, a well-known dilemma that innovators face, is not only affected by, but also change knowledge space where innovations occur. To investigate how the trade-off contributes to knowledge space, we present a network model that sub-linearly grows by following Heaps' law and creates links with adjacent possibles, and compare model networks to empirical knowledge spaces constructed from patents and research papers. With the balanced trade-off, the model reproduces structural properties found across empirical knowledge spaces, including broad strength distribution, significant local clustering, and especially modular structure that is often considered taken for granted. If either exploitation or exploration dominates the trade-off, these structures could not be obtained. Interestingly, the empirical spaces appear to evolve with the balanced trade-off, suggesting collective behaviors that balance exploitation and exploration naturally shape knowledge modules. Our network model not only explains why modular structure exists within knowledge spaces but also provides insights into the co-evolution of microscopic behaviors and modular structure.
45 |
46 |
47 |
48 |
49 |
50 |
51 |
Friday, 03/06
52 |
11:00am-12:20pm
53 |
1155 E 60TH ST, Room 289A
54 |
55 |
56 |
57 |
A light reception will be provided by Northern Taste following the talk. The reception will be held in the new MACSS lounge, 1155 E 60TH ST, Room 224.
58 |
59 |
60 |
61 |
62 |
63 | **Hyejin Youn** is an Assistant Professor of Management & Organization Department at the Kellogg School of Management, and a core faculty at NICO, the Northwestern Institute on Complex Systems. She is also Royal Society of Arts fellow, and an external fellow at London Mathematical Laboratory, London, UK. Prior to joining Kellogg, she worked at University of Oxford, Harvard University, and MIT Media Lab, and Santa Fe Institute, as a research fellow. Hyejin received her PhD in Physics in 2011 from Korea Advanced Institute of Science and Technology (KAIST). She was a Principal Investigator of the project a National Science Foundation grant (USA) to study Technological Change from the Map of Capabilities.
64 |
65 | Her research interests are to understand the interplay between technological innovation and socio-economic systems (urbanisation, economic diversity and specialisation, invention activity, future of work). Her highly interdisciplinary approach often results in broad collaborations ranging from mathematicians, computer scientists, economists, sociologists, anthropologists, to archeologists. Her work has been published in general audience such as Nature communication, and PNAS, as well as top specialized journals such Physics Review Letter, and Evolutionary Anthropology, and has been featured in The Econonmist, Forbes, The Guardian, WIRED, Scientific America, MIT Technonlogy Review, among other major global media outlets. Her goal is to develop a theoretical, yet empirically grounded, framework that will enable us to turn the increasing volumes of data into scientific insights and well-designed policies, an approach known as computational social science. The mathematical tools and computational methods that are used include scaling theory, spatial analysis (including percolation theory, information theory and fractal dimension analysis), statistics, and network theory.
66 |
67 |
68 |
69 |
70 |
71 | This week's suggested readings:
72 |
73 | - [Van Der Wouden, Frank and Hyejin Youn. 2020. "Impact of geographical distance on acquiring know-how through scientific collaboration." Working Paper.](https://github.com/uchicago-computation-workshop/Winter2020/blob/master/03-06_Youn/youn_know-how.pdf)
74 |
75 |
76 |
77 |
78 |
79 | ---
80 |
81 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
83 |
84 | **Summary:** During the last ten years, network science has emerged as a strongly
85 | interdisciplinary new field of studies in academia. Here, I will trace
86 | the epistemic development of the field examining the mathematical
87 | origins of certain important concepts of network analysis as they were
88 | exported mainly from the rudiments of graph theory. I will discuss a
89 | few examples of schools of mathematics (mostly in USA) starting to
90 | offer courses and to host research on network science. Moreover, I
91 | will present the Wikipedia network stemming from the 3 Wikipedia pages
92 | (Graph theory, Network science, Complex network) and I will discuss
93 | the structure of this Wikipedia network based on the graph-theoretic
94 | notion of minimum dominating sets. The exemplary network that will be
95 | presented here is that of the (direct and) indirect collaborations
96 | among members of the NYUAD Faculty of Mathematics generated by data
97 | extracted from the MathSciNet database of the American Mathematical
98 | Society. Furthermore, I will analyze the bipartite graph of faculty
99 | members and research interests using the tools of Formal Concept
100 | Analysis. Finally, I will model the effect of a network influence
101 | process on the Math collaboration network when research areas are
102 | sources of boundary influence simulations.
103 |
104 |
105 |
106 |
107 |
108 |
109 |
Thursday, `r date`
110 |
11:00am-12:20pm
111 |
Cummings Life Science Center, Room 101
112 |
113 |
114 |
115 |
116 |
117 |
118 |
119 |
120 |
121 | **`r speaker`** is in the Faculty of Northwestern University School of Professional Studies Data Science Program. Currently he is Visiting Professor of Mathematics at the New York University Abu Dhabi (NYUAD), where he is affiliated to the Research Group in Network Science [RGNS](https://sites.google.com/nyu.edu/rgns/home).
122 |
The 2019-2020 Computational Social Science Workshop meets each Thursday from 11 a.m. to 12:20 p.m. in the Cummings Life Science Center, Room 101. All interested faculty and graduate students are welcome.
31 |
32 |
33 | **Dashun Wang** - Associate Professor of Management and Organizations at the Kellogg School of Management
34 |
35 |
36 |
37 |
38 | ### January 30th - [Joshua Becker](https://github.com/uchicago-computation-workshop/Winter2020/tree/master/)
39 |
40 |
41 |
42 |
43 | **Joshua Becker** - Postdoctoral Fellow, Kellogg School of Management, Northwestern Institute on Complex Systems
44 |
45 |
46 |
47 |
48 | ### February 6th - [Andrei Boutyline](https://github.com/uchicago-computation-workshop/Winter2020/tree/master/02-06_Boutyline)
49 |
50 |
51 |
52 |
53 | **Andrei Boutyline** - Assistant Professor of Sociology at the University of Michigan
54 |
55 |
56 |
57 |
58 | ### February 13th - [No workshop this week]()
59 |
60 |
61 |
62 | ### February 20th - [No workshop this week]()
63 |
64 |
65 |
66 |
67 | ### February 27th - [Diogo Ferrari](https://github.com/uchicago-computation-workshop/Winter2020/tree/master/02-27_Ferrari)
68 |
69 |
70 |
71 |
72 | **Diogo Ferrari** - Assistant Instructional Professor in Computational Social Science at the University of Chicago
73 |
74 |
75 |
76 |
77 | ### March 6th - [Hyejin Youn](https://github.com/uchicago-computation-workshop/Winter2020/tree/master/03-06_Youn)
78 |
79 |
