Veranstaltungen

• Introduction to Crowd Innovation Lab at Harvard, the Lab with a mission to study the contests, the crowds, and to use the crowds to crack challenges for NASA and Academia

• The insights on crowdsourcing - advantages, trade-offs and niches

• Algorithm and Data Science challenges as a sweet-spot of crowdsourcing - examples, results and stories

Dr. Rinat Sergeev is Senior Data Scientist & Chief Scientific Advisor at the Crowd Innovation Lab/NASA Tournament Lab at Harvard University. Rinat works as a head of data science team, and a lead science and technical expert on exploring and utilizing crowdsourcing approaches in application to the data science and algorithmic challenges, coming from NASA, Business, or Academia. Rinat received his PhD in Quantum Mechanics in Ioffe Institute, Saint Petersburg. His research interests include conceptual analysis, analytical approaches and models in multiple areas.

Abstract:
This natural field experiment tests the effects of purely symbolic awards on volunteer retention in a public goods context. The experiment is conducted at Wikipedia, which faces declining editor retention rates, particularly among newcomers. Randomization assures that award receipt is orthogonal to previous performance. The analysis reveals that awards have a sizeable effect on newcomer retention, which persists over the four quarters following the initial intervention. This is noteworthy for indicating that awards for volunteers can be effective even if they have no impact on the volunteers’ future career opportunities. The awards are purely symbolic, and the status increment they produce is limited to the recipients’ pseudonymous online identities in a community they have just recently joined. The results can be explained by enhanced self-identification with the community, but they are also in line with recent findings on the role of status and reputation, recognition, and evaluation potential in online communities.

Abstract:

This paper examines patent oppositions as firm-level responses to newly-granted patents. We present a citation-based construct for measuring the technological lineage to which a newly granted patent lays claims. This network-analytic construct, which we label technology trees, allows us to develop a refined explanation of patent oppositions by taking into account the ownership structures of the technology to which a particular patent relates. An application to the cosmetics industry reveals that the technology tree measure, which is calculated on the level of individual patents, usefully complements established network measures at the industry level, such as triplet counts for measuring patent thickets. This suggests a need for further and more fine grained analyses of technology trees as context in which patent oppositions play out.

Abstract:

We study how competitive mechanisms of external knowledge acquisition affect the emergence of collaborative modes of learning. Specifically, we link the hiring of R&D scientists from competing firms to the subsequent formation of collaborative agreements between these firms – in our setting technology oriented alliances and licensing agreements. Using data on inventor mobility and alliance formation amongst 42 global pharmaceutical over 16 years we are able to show that inventor mobility in fact increases the likelihood of alliance formation in periods following inventor moves. Moreover, our results suggest that this effect becomes more pronounced if more productive inventors are recruited but weaker if the recruiting firm is already familiar with the inventor’s prior employer’s technological capabilities. These findings are more robust for alliances that aim at joint R&D efforts than for pure licensing agreements.

Please drop us a line if you plan to attend.

Abstract
We study bilateral cross-licensing agreements among N (> 2) competing firms. We find that the industry-profit-maximizing royalty can be sustained as the outcome of bilaterally efficient agreements. This holds regardless of whether agreements are public or private and whether firms compete in quantities or prices. We extend this monopolization result to a general class of two-stage games in which firms bilaterally agree in the first stage to make each other payments that depend on their second-stage non-cooperative actions. Policy implications regarding the antitrust treatment of cross-licensing agreements are derived.

Abstract
Using a large scale experiment involving more than a thousand open source software programmers, we match individual monthly contributions to open source programs with behavior in online games we ran. We study how social preferences affect the patterns of production and the overall success of projects. We show in particular that groups with a larger share of reciprocators are more likely to fail, but conditional on not failing, are more successful. The effect is particularly strong for technological areas and development stages characterized by more variance in contributions.