Mnikr: reputation construction through human trading of distributed social identities

Citation: Brendan Francis O'Connor, John Linwood Griffin Mnikr: reputation construction through human trading of distributed social identities.
DOI (original publisher): 10.1145/1655028.1655032
Semantic Scholar (metadata): 10.1145/1655028.1655032
Sci-Hub (fulltext): 10.1145/1655028.1655032
Internet Archive Scholar (search for fulltext): Mnikr: reputation construction through human trading of distributed social identities
Wikidata (metadata): Q65761353
Download: http://jlg.name/papers/2009-11-13-dim09-mnikr-reputation-construction-through-human-trading-of-distributed-social-identities.pdf
Tagged:

Summary

RQ1: Can the idea of buying and selling shares on a theoretical market be used to gain a machine-readable understanding of traditionally unquantifiable data, without resorting to the techniques of machine learning or artificial neural networks?

If RQ1 is true:

RQ2: How effective would such a system be at establishing measurably different reputations?

RQ3: How well would its reputations correlate with real-world perceptions of reputation?

RQ4: Would such a system truly be useful across multiple fields, rather than having applicability only to one context?

To answer RQs, designed and implemented Mnikr, a system that "automatically generates combined digital identities, using publicly-available information sources to find both different parts of one identity, and the links between different identities. Users of Mnikr can then buy and sell shares of these identities on a stock exchange, and this process determines the value of an identity’s reputation."

Ran Mnkir for 32 days on the internet. 41 users registered, and the system collected 351,356 identities, comprised of 543,902 validated personas. Showed some level of promise toward RQs, eg high scoring reputations were for famous programmers.