WWN Features Compared with Related Products

WWN is a network technology that allows network users to find valuable but sensitive information while minimizing the risk that the sensitive data is misused or misdirected. There are no current network products that achieve this goal, although there is a new product from IBM called Anonymous Resolution or AR that makes similar claims. On this page, we will highlight the product features by comparing WWN with AR and also with an infrastructure product that is an essential companion to both, Public Key Encryption Infrastructure (PKI).

To achieve the goal, a network system needs a variety of features. These are listed in the following table that also indicates which products have the features.

WWN may be new in the world of computer networks, but it is based on a published, and well-supported conceptual framework. Let us describe this framework and then compare the features of WWN and AR using the previous table as an outline.

Features Needed for Negotiating a Win-Win Cooperation

Our view of negotiation is rooted in advances made in game-theory during the 1980's. This work lead to important insights into games that display aspects of the "Prisoner's Dilemma". Briefly stated, the prisoner's dilemma exists when there is an expected payoff for a win-win cooperation with another party but the expected payoff is even larger for betraying the other party and forcing a win-lose outcome.

Such situations abound. Before we negotiate, the first question should be "What can I lose"? If the risk is acceptable, the second question is what strategy to adopt in the face of the prisoner's dilemma Finally, if the game involves the network, you must ask whether the network enables you to adopt the preferred strategy. Currently, the answer to that final question is "no". That is why WWN technology is needed for future Internet systems.

Because current network technology makes it too easy to betray trust, responsible organizations have erected security barriers around their data. These barriers prevent cooperation -- even cooperation that might lead to win-win results for all involved. To end this data blockade, we need to take into account the key conclusions of game-theory research:

1. Know who you are dealing with and how they have behaved in the past.
2. Reward negotiation partners who aim for win-win and punish those who try to win at your expense (Tit-for-Tat strategy)
3. Play the game for small stakes on each transaction to avoid a big loss if betrayed.
4. Keep options to break-off negotiation open as long as possible.

This research is not purely theoretical, by the way. It is grounded in extensive experimentation and a careful analysis of human behavior in different societies and throughout history. Broad principles emerge forcefully from the data and it is high time to put the principles to work in computer network systems. Now let's take a look at two products that claim to do that.

Anonymous Resolution from IBM

In 2006, the IBM corporation brought to market a groundbreaking new product called Anonymous Resolution (AR). It is so new, we will provide a capsule description here. Our remarks are based on material from the IBM web site.

IBM describes AR in these words "It enables multiple organizations to compare proprietary data in a manner that identifies relationships and develops leads but never exposes sensitive data values." The documentation ascribes the novel capabilities of AR to the one-way hash functions employed to encrypt names and other identifiers. With this system, it is safe to expose the encrypted data on the Internet where it is unreadable by unauthorized users. Authorized organizations that have a copy of the encryption key and can compare encrypted values to find matches in the data.

AR clearly presumes PKI so that only authorized organizations receive the encryption key. It also goes a long way towards protecting the data by preventing attacks from unauthorized Internet users. On the other hand, AR remains a basic client-server system familiar to all Internet and corporate data users. A simple client-server is powerless to prevent an authorized - but unethical - user from exploiting the data for profit at the expense of the data owner. That brings us back to the Prisoner's Dilemma. If you want to keep people from taking advantage of you, you can't put all your valuable data on the Internet -- even if you have AR in place.

WWN Features for Win-Win Cooperation

The main feature of WWN that distinguishes it from AR is a three-part, brokered architecture consisting of network sites for all parties in the negotiation and a common site for a party acting as a broker. Each site plays both client and server roles at different points in the negotiation.

Negotiation occurs between pairs of parties leading to information exchange facilitated by the broker. The third party in a negotiation is a broker who identifies small pieces of matching information held by the two other parties. If the negotiation proceeds to conclusion, only the matching pieces are exchanged. The key invention of WWN is that the broker cannot actually read the information; therefore, the broker cannot compromise the data security. The parties in the negotiation should agree to exchange only the matching data. That limits the liabilities of both to the small amount of content that is exchanged.

None of the individual parties in the WWN negotiation has a sufficient set of encryption keys to compromise the system. The only vulnerability in WWN, and this cannot be ignored, is that a broker can secretly team with one of the other parties to betray a third party. For that matter, the provider of the encryption keys could team with one of the parties to subvert the security. Although there is no technical solution to every manner of unethical behavior, the WWN technology does dramatically reduce the risks associated with simple client-server information systems.

Here is a list of the important features of WWN that encourage win-win cooperation:

1. Only small amounts of concealed information are ever transferred between parties.
2. Negotiating positions can be tailored to particular negotiation parties. This feature can be employed to reward reliable partners and punish selfish parties. The feature is well advertised so it encourages good behavior.
3. Parties can refuse to cooperate with particular partners and this fact can be concealed along with other aspects of the negotiating position.

The overall effect of WWN is ideal for encouraging principled cooperation because it forces a series of small negotiations. Between negotiations, each party can review the benefits and problems arising from the behavior of the other parties. During subsequent negotiations, the degree of cooperation can be made contingent on good behavior.

Features that Improve Matching