Monday, August 08, 2005

Human Markup Language (Human ML)

Information technology (IT) that mankind devises is merely means to accomplish the task of sharing their thinking, coordinate, work together to accomplish things, leave a legacy of knowledge and wisdom gained, entertain each other, and socialize.Human markup language (HumanML) is one possible technology for addressing some of these critical aspects of communication by providing machine processable subtext through the use of extensible markup language (XML).
HumanML is a new specification being developed by the Organization for the Advancement of Structured Information Standards (OASIS).

"Some of the questions that the use of HumanML could hopefully address are: Is the retrieved document informational in nature? Is it intended as policy, as advertisement, propaganda or some other purpose? At the point in history of document creation, what was the unstated motivation to create and publish it? What was the author's attitude toward the subject? When extracting or quoting short sections for citation or comment, how can this context be carried forward without the need to include long background passages? How to avoid accidental release of protected information by ignoring or forgetting to include external markings? How to avoid distortion of original meaning or intent whether intentional or not? How to understand cross-culturally and interpret specialized lingo such as "governmentese" in current layman's terms and in the current context?"

Goals of HML:
  1. Providing a standardized means to convey and establish contextual meaning is intended to allow authors a chance to rise above the chaos described below and permit researchers more opportunity to timely pull valuable nuggets of information and knowledge out of that same chaos.
  2. Another goal of HumanML is also to provide document markup standards that permit adjustment of the human computer interface (HCI) to the system users' most effective modality.

Author clearly says, "beyond etymological meaning of words that dictionaries and related technologies provide is semantic meaning of words in their context - something addressed by natural language processing (NLP) technology and the Semantic Web, which are just now beginning to emerge from the domain of cutting edge research. It is in this environment that HumanML is to find its niche."HumanML complements other semantic based technologies by providing for inclusion of human related context in the form of standardized markup.

A must read article HumanML: The Vision. Just after I reading it I remembered the state of blogosphere post at http://www.sifry.com/alerts/ where the statistics clearly show that one blog is created every second. And context based mining as used in yahoo search. Hoping for some context based and symantic analysis soon in other search engines.

Assuming most of people know about PMML,(Predictive Model Markup Language),
http://www.dmg.org/, which is an XML mark up language to describe statistical and data mining models.

I think sometime these two and blogosphere will combine to give better results on text mining and helping the internet to bring people more closer to each other by better symantic analysis using human pschology ontologies, text mining algorithms (semantic analysis and statistics).

0

0 Comments:

Post a Comment

<< Home