The History Behind Broadcatch

Social Advertising (1994)

Abstract

Social Advertising uses broadcatch agents to facilitate resource discovery & use, building trusted communities through personal and "market-driven" selection with authenticated reputations and shared interests as a basis.

History

Current research in networked information access is making great strides in providing the ability to create personal, shareable views onto the universe of information available on the Internet. One of the major problems confronting each of these systems is that of resource discovery: how do you find something if you don't know here to look for it?

WAIS (Kahle & Medlar, 1991) offers indexes, but the indexes must first be located (through a hierarchy of server directories), and secondly the manner in which the files are indexed may not be suitable to every conceivable task.

Prospero (Neuman, 1992) and the World Wide Web (CERN, ?) give personalized views that may be shared, so you can choose your own "expert editors" for each field that you are interested in, and even become one yourself. But finding the experts in the first place is, again, a bit of magic.

Tapestry (Goldberg et al., 1992) allows you direct access to human collaborators that can aid you in your filtering, but the system is designed mainly for filtering information as it comes in -- less so for user initiated searches.

What all these systems can benefit from is a mechanism for advertising questions as well as sources of answers.

Scenario

Imagine sitting at your Mosaic client, and you want to find out what the exact dimensions of an official Ultimate Frisbee playing field are. It might take a while through standard means to find this -- in fact, you might end up giving up before finding the URL

   "https://www.cs.rochester.edu/u/ferguson/ultimate-rules.html#III"

One common fallback mechanism is to post your request in a related Usenet conference. Unfortunately, there is no frisbee news group, and you might not know that aficionados generally prefer the more generic term disc and thus overlook rec.sport.disc.

So you give up and post your request in rec.games.misc or, more probably, in some totally unrelated newsgroup but one in which you feel comfortable along with an apology for the off-topic request.

Social Advertising

With Social Advertising, the scenario is a bit different. Again, let's start with the Mosaic client, but now we do something new: we open a page with the Social Advertising request form, and type in our request:

   Social Advertising: Request Page
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   KEYWORDS:  games, outdoor, ultimate frisbee, rules
   REQUEST:   what is the size of the playing field?

This request gets posted on a virtual bulletin board, along with your return address. The keywords line is not required, but it helps automatic agents parse and file your request where it can be handled most directly.

If someone has posted a resource page including, say, ultimate frisbee, then there could be a direct match right away, and the URL would be returned to you automatically. Such a page may look like this:

   Social Advertising: Resource Page
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   KEYWORDS:  games, ultimate frisbee, rules
   TITLE:     UPA Rules of Ultimate, 9th Edition
   URL:       https:// ...path elided... /ultimate-rules.html

However, if there were no such match, the request would be orphaned, and there would exist services (and people) who would be interested in finding results for orphan pages.

Building a Community

One of the goals of Social Advertising is to aid in the building of virtual communities. So far, we have described a mechanism wherein people can help each other find things, but why would people choose to do so? Or, more specifically, what's in it for them?

Here are three answers to this question, but we believe there are more:

1. Advertising one's resources is generally advantageous.
2. As you help others who have similar interests, you can keep track of them and perhaps use them as a resource at a future date.
3. With the advent of a digital credit (or cash) system, people will be able to get paid for the information they provide, or even the meta-information of directing others where to look.

Note that (2) is a community builder. Both sides increase their sum of information at this point, as each learn of a source of potential future collaboration in similar areas (or, if the information is bad, a source perhaps to ignore in the future!). This is where cooperation begins to come into play. But first, a brief statement regarding personal privacy.

Privacy

In the scenario above, we assume that return addresses are available for all parties, as the agents must know where to return found information, and as part of the cooperative step, provide a handle upon the source for future use. But as the set of questions or resources one advertises enlarges, profiles can be built that compromise one's privacy. Further, there are some questions one simply may not ask for fear of reprisal.

The Social Advertising process depends upon authenticatable pseudonym servers and anonymous bulletin boards for the provision of a high degree of personal privacy. More on this is outside the bounds of this introduction.

Filtering

Personalized fitness functions and market driven selection mechanisms help the user build a trusted community of resource discovery systems. These filters may themselves be discovered through resources found through Social Advertising.

Conclusion

We believe that access to information in a manner determined by the accessor and not by the provider, along with the ability to communicate and share information without fear of reprisal are essential elements of a free and open society, and the first steps towards a well-educated and compassionate community.

Social Advertising puts the power of information access into the hands of the user.