Memex and the Personal Data Economy
Wikipedia’s defines memex as a portmanteau of memory and extension, a term coined by Vannevar Bush in his influential article titled As We May Think. In layman terms, memex means a personal data store that contains stuffs that matter to a particular user.
In our daily computing activities, we generate data. From Amazon.com wish lists and purchase receipts to Facebook contacts and status updates. From Delicious bookmarks to Twitter tweets. From personal finance activities to browser clickstream history. We create and store certain types of data such as bookmarks with the explicit intent of using them at a later time, whereas other types of data such as browser clickstream history may appear to have lesser values to some (and thus get purged from time to time).
Looking from across the table with the perspective of service providers, our personal data means cash. Amazon.com uses our past purchases to recommend new products. Mining browser clickstream to create marketing segments is at the heart of the fast growing behavioral targeting (BT) industry. Finally, in the lucrative battlefield of search, if the ultimate goal is to computationally understand user intent, or as Sergey Brin put it, connect search straight to the brain, then it is necessary to go beyond the current search query and construct a deep, stateful understanding of users based on, yes, their personal data.
The above are examples of personal data driven applications, henceforth referred to simply as memex agents or just agents, that will become increasingly more powerful and demanding increasingly more data. Yet these applications face two fundamental problems: data silo and privacy. Data silo manifests for example in the fact that our search engines do not see our Amazon wish list, and Amazon does not see our search history. Consumer privacy concerns have recently been an increasingly important and controversial topic that I will not dwell into here, except to point out that they are often exacerbated by businesses’ attempts to break the data silo, especially in BT: personal clickstream, demographic profiles, and even offline data are being bought by companies trying to expand their data portfolio for improved personalization and targeting.
I believe that there is an excellent opportunity to build an open user-centered data platform that comprehensively addresses data silo and privacy issues, and thus catalyzes dramatic improvement in agent applications. A schematic illustration of this platform looks like below:
The key idea is to empower the user with tools to take control of and be the integration point of her data. Each user will have the opportunity to create her own personal data store that lives across “three screens and the cloud”. This virtual data store is the user’s holistic memory extension that stitches together the user’s currently fragmented online identities across Facebook, Twitter, Bing/Google, Delicious, Amazon, etc. In this user-centered world, the silo problem will apparently no longer exist, and the privacy problem will now be transformed into one of control and fair value exchange. User control means that the user has complete, email-like control over her memex data. Fair value exchange means that an agent that requests access to the user’s memex data will have to clearly explain its quid-pro-quo (which can run the gamut from personal informatics utilities to access to premium content to the promise of improved advertising relevance).
The dramatic improvement in memex agents stems from the potential of breaking down the silo and mashing up personal data. Think about Google-aware Amazon recommendation, or Amazon’s wishlist-aware Bing shopping experience. We don’t want seams across our three screens and the cloud. We can certainly benefit from having no seam across our online identities.
Whoever succeeds in building such platform will be the winner in the resulting personal data economy, which Forrester refers to as the personal cloud.
In the upcoming posts, I will write more about requirements for the memex platform, as well as the companies that are candidates to build it. Stay tuned.
November 1, 2009
Posted in: Uncategorized
3 Responses
Tweets that mention Dreaming Engineer В» Memex and the Personal Data Economy -- Topsy.com - November 1, 2009
[...] This post was mentioned on Twitter by Greg Lloyd, Vu Ha. Vu Ha said: 1st real blog post: Memex and the Personal Data Economy http://bit.ly/CTW40 [...]
Personal Cloud: A Microsoft Take « Reverend Ted’s Blog - November 4, 2009
[...] I recommend checking out his post. [...]
Dreaming Engineer » Memex: Perspectives From The User - November 9, 2009
[...] an earlier post, I wrote about the opportunity to create a user-centered personal data platform called memex. Memex [...]
Leave a Reply