3 ms·

Docs Are Old-School, We Need PageRank for People

24 points by dnewcome 17y ago

leecho0 17y ago
I think he brings up a good point, but I have serious doubts about google using it because of privacy issues. What he suggests is that there is a ranking for each person in the world, and the content he creates will be ranked higher in the world. Merely tracking the same author could either be an invasion of privacy -- tracking the ip and/or taking a peek at the person's cookie, or ridiculously hard -- analyzing usernames, vocabulary, tone of voice, etc. He mentioned doing something special with openid, but it would need widespread adoption before this could be done.
old-gregg 17y ago
"PeopleRank" he's suggesting is good for jut two very special kinds of individuals. First group is public figures and marketing people. Visibility and being relevant means everything to them. But even these folks often maintain dual personality life i.e. public/true versions of themselves. Some smarter programming contractors also fall into this category, although they prefer to call themselves "consultants" presumably because it lacks an unwanted association with plumbers, roofers and other contract workers of "lesser glamour". The second group is pretty much "first group wannabes", i.e. highly social individuals crying for more attention. Most blogging programmers fall into this category, i.e. folks who (instead of actually coding) can spend an evening writing a post on something as pointless/dumb as "RDBMS is obsolete" prettying it up with irrelevant photos of random objects. No, they aren't any more "experts in AJAX" than folks who don't blog, they just want attention. But most of us don't belong to either of these two groups. When I attended the startup school last year I met about 50 different HN folks, and most of them have never posted anything here, let alone using their real name. "PeopleRank" would only anger them, I suppose.
- pudo 17y ago
  While you are probably right in identifying these two groups as very intense users of such a system, I would argue that they would not be its typical users but rather the people most likely to abuse it. This has to do with the concept of "personal online reputation", which is often used as a weasel word for "stuff people will find when they google you". This, of course, is something you can easily influence and leads to a form of digital masturbation that I am highly critical of. If, otoh, you use a more scientific definition, e.g. "the aggregate of all opinions of other people that have interacted with your previously and that are available in a given social network", this becomes a very interesting measure. And one that is incredibly hard to work with. But, assuming that someone figured out how to handle it, it would be incredibly useful: it would allow for ad-hoc trustful pseudonymous interactions both in situations like the current struggle in Iran, where certain pseudonyms could quickly gather domain-specific reputation and thus become effective gate keepers. The value of such a mechanism extends through all forms of online cooperative services and could be extremely valuable in fighting trolls, griefers etc. Another potential use of this is around privacy: While I believe that we need to find many new (social and legal) norms regarding the handling of personally identifying information in the web, I also think that we will have to develop a generation of post-privacy tools that will help us deal with data that has leaked and will never be re-secured. Creating multi-faceted, rich online identities might be such a tool. These scenarios will be interesting to more users than just marketing professionals and wannabes, although it is likely that the new discipline of reputation corruption will grow right alongside with these tools.
natch 17y ago
This is a very old idea. It's been given various names: FriendRank, PeopleRank, PersonRank -- many, many hits on Google.
- 10ren 17y ago
  Also "web of trust"
  - twopoint718 17y ago
    I was thinking the same thing. Having a chain of signed PGP keys that allow you to infer the trustworthiness of someone hasn't seemed to catch on. Are there any good articles on how this idea developed, and succeeded and or failed (I'm sure it was/is in very heavy use somewhere or among some group)? Heck I once went to a signing party... http://en.wikipedia.org/wiki/Key_signing_party http://en.wikipedia.org/wiki/Key_signing_party
- Marshall_Clark 17y ago
  Those just have similar names. It's actually a new combination of a bunch of established ideas: 1) PeopleRank 2) PageRank 3) Hilltop 5) Semantic Analysis 4) Web of Trust 5) Consumer Authentication Services

24 points by dnewcome 17y ago

I think he brings up a good point, but I have serious doubts about google using it because of privacy issues. What he suggests is that there is a ranking for each person in the world, and the content he creates will be ranked higher in the world. Merely tracking the same author could either be an invasion of privacy -- tracking the ip and/or taking a peek at the person's cookie, or ridiculously hard -- analyzing usernames, vocabulary, tone of voice, etc. He mentioned doing something special with openid, but it would need widespread adoption before this could be done.

old-gregg 17y ago

"PeopleRank" he's suggesting is good for jut two very special kinds of individuals. First group is public figures and marketing people. Visibility and being relevant means everything to them. But even these folks often maintain dual personality life i.e. public/true versions of themselves. Some smarter programming contractors also fall into this category, although they prefer to call themselves "consultants" presumably because it lacks an unwanted association with plumbers, roofers and other contract workers of "lesser glamour". The second group is pretty much "first group wannabes", i.e. highly social individuals crying for more attention. Most blogging programmers fall into this category, i.e. folks who (instead of actually coding) can spend an evening writing a post on something as pointless/dumb as "RDBMS is obsolete" prettying it up with irrelevant photos of random objects. No, they aren't any more "experts in AJAX" than folks who don't blog, they just want attention. But most of us don't belong to either of these two groups. When I attended the startup school last year I met about 50 different HN folks, and most of them have never posted anything here, let alone using their real name. "PeopleRank" would only anger them, I suppose.

pudo 17y ago
While you are probably right in identifying these two groups as very intense users of such a system, I would argue that they would not be its typical users but rather the people most likely to abuse it. This has to do with the concept of "personal online reputation", which is often used as a weasel word for "stuff people will find when they google you". This, of course, is something you can easily influence and leads to a form of digital masturbation that I am highly critical of. If, otoh, you use a more scientific definition, e.g. "the aggregate of all opinions of other people that have interacted with your previously and that are available in a given social network", this becomes a very interesting measure. And one that is incredibly hard to work with. But, assuming that someone figured out how to handle it, it would be incredibly useful: it would allow for ad-hoc trustful pseudonymous interactions both in situations like the current struggle in Iran, where certain pseudonyms could quickly gather domain-specific reputation and thus become effective gate keepers. The value of such a mechanism extends through all forms of online cooperative services and could be extremely valuable in fighting trolls, griefers etc. Another potential use of this is around privacy: While I believe that we need to find many new (social and legal) norms regarding the handling of personally identifying information in the web, I also think that we will have to develop a generation of post-privacy tools that will help us deal with data that has leaked and will never be re-secured. Creating multi-faceted, rich online identities might be such a tool. These scenarios will be interesting to more users than just marketing professionals and wannabes, although it is likely that the new discipline of reputation corruption will grow right alongside with these tools.

natch 17y ago

This is a very old idea. It's been given various names: FriendRank, PeopleRank, PersonRank -- many, many hits on Google.

10ren 17y ago
Also "web of trust"
- twopoint718 17y ago
  I was thinking the same thing. Having a chain of signed PGP keys that allow you to infer the trustworthiness of someone hasn't seemed to catch on. Are there any good articles on how this idea developed, and succeeded and or failed (I'm sure it was/is in very heavy use somewhere or among some group)? Heck I once went to a signing party... http://en.wikipedia.org/wiki/Key_signing_party http://en.wikipedia.org/wiki/Key_signing_party
Marshall_Clark 17y ago
Those just have similar names. It's actually a new combination of a bunch of established ideas: 1) PeopleRank 2) PageRank 3) Hilltop 5) Semantic Analysis 4) Web of Trust 5) Consumer Authentication Services