December 2007 - Posts - Dreams Come True

Tag, you're it

There’s a lot of talk across the web these days about tags, tagging, tag clouds, tag services. What exactly is a tag? How would I explain it to my five year old daughter [who even though wicked smart, isn’t going to understand the semantic web lingo I would use to explain tags]?

At a crude level, I like to think of tags as a way for me to find stuff, then categorize it with a term I have a relationship with [the tag is part of my vocabulary]; thus creating a relationship between the term relationship and the stuff I find [did that make sense?].

Now what I have yet to see explode on the web is tags which cross spoken languages; which cross term [if you will] libraries. This shouldn’t be confused with tag aggregation, which the big services [del.icio.us for example absorbs all tags – from my perspective – in bulk load appetites] do fairly well. Could tagging be the perfect hybrid of translation and cross language search? Imagine a scenario where one industry group could ‘talk’ to another industry group via tag crossing.

There’s also tag authoring limitations; especially from a mobile device. As good as the G folks are at simplifying mobile services, they completely miss the mark on features like tag management. Is a cell phone number a semantic expression for the person who answers on the other end; therefore we should be able to tag the number to web-ify it? Where is this feature [from the new open cellular platforms, and their great ambitions]? More on all this later.

Con paz,

frank

Posted Dec 18 2007, 02:46 PM by frank martinez with no comments

ID Dust

We took a break last week from the white hot heat of building a web platform, to check out "Golden Compass". The timing was interesting for me because I happened to be reading "Galileo's Daughter". In both stories science combats control imposed by a large body of [percieved] powerful people, and their fear of distributing knowledge.

We are fortunate at Microsoft Research to explore new worlds without constraints imposed by a large body of [fill in the blank, depending on where you might have encountered opposition to new ideas]. Instead, distribution of knowledge and collaboration with communities interested in this knowledge is expected in all facets of our work at MSR.

But let's bring it back to the web. I've been thinking about control as it relates to any person who uses the web [for the sake of this blog post, will defer 'internet' to transport supporting the web]. How much control does any person have on the web? And this gets more complicated as the web strives too become conversational, or transactional. APIs are talking to other APIs; do humans have a say in all this cross platform chatter? Have they been marginalized by a large faceless [not to be confused with facebook :> ] body making decisions in real time about what each person wants.

If I can ask for an industry shifting gift [during this holiday season], it would be to return control to those who use the web everyday. Give them the power to control their identity, their read/write operations against any platform, their data collected from searches and other discovery activity. What would the web look like if this happened? How would web platforms need to evolve, to support power returned to those who use the web?

Demo Day

One of the great joys of a PM is to gather an audience into a room, fire up a laptop, point it at a screen - say a quick prayer - then unveil an idea come to life. It's what you live for in this [software] business, why I love working at Microsoft and especially Microsoft Research. Not all demos go well; sometimes the ideas are ahead of their time or not enough time was available for fit/finish, and the idea gets lost in translation. But when a demo does go well all the digital stars seem to align in one magic moment.

In the coming weeks we'll be doing some demos, get the feedback loop started on a few ideas we've been incubating all summer. I'll update the community when there's internet facing bits we want your feedback on.

Thanks to all who made our demo days perfect, no matter what happens.

frank

Posted Dec 06 2007, 01:18 AM by frank martinez with no comments

Light and Fast

Hi,

I'm the Lead Program Manager on the team building the next generation of research.microsoft.com and related services. We decided months ago [long before I arrived] to re-think what a web presence should be for Microsoft Research. Building a web 'site' isn't necessarily difficult. In contrast, developing a conversational platform to distribute assets across web and other end points is an exciting journey. I use the word conversational to suggest there is dialog between people interested with Microsoft Research - they use the web to maintain this interest - and the platform which feeds their appetite for more information.

There are other terms [or phrases] to suggest what we're up to: semantic, read/write, 2.0, service oriented. Yet none of these terms suit the problem we're faced with - distributing information as it's desired in the context for which it will have the most meaning and relevance. With the next generation of research.microsoft.com we intend to explore new ways to match the intent of 'people interested with Microsoft Research' with a response [from the platform] to complete [or perhaps get started] the conversation.

con paz,

frank

Posted Dec 04 2007, 11:50 PM by frank martinez with no comments

Microsoft Research Community

Dreams Come True