Intent of Work
Teknowledge Corporation, Palo Alto
March 19, 2001
1. Technical Goal/Accomplishment
We will provide Agent Semantic Communications Service - a DAML enabled search and translation service capable of coping with the semantic heterogeneity of the web. The system will be highly distributed and will scale to the size of the web of both today and tomorrow. It will have performance equivalent to today's text-based search engines. ASCS will be easier to use than today's search engines because it will search for content that has the necessary meaning not just surface syntax to answer the query posed to it. ASCS will have a clearer interface than today's search engines because it will support queries similar to human language rather than the ambiguous state of character-based queries which may either be words, phrases, character strings or sometime expressions like database query languages.
2. Potential Users
The potential end users of our technology include anybody who has the need to use semantically interoperable information agents on the web. For example, military commanders can collect intelligence reports from different sources and make battlefield situation assessment. Intelligence analysts will be able to obtain higher precision and recall in their search. Scientific research community will be able to share on the web the data collected in different labs for different purposes. A bank can gather more complete and accurate financial information about a customer before a loan application is approved. As a potential transition path to the financial community, we plan to bundle DAML with our TekPortal, an existing commercial product for financial information portal. This will make the semantic content of financial reports available to partners without the need for tedious levels of agreement on database fields and formats. We are also developing an ontology of financial information that also covers the semantic content of XML-based standards like OFX. This will ensure the feasibility of moving financial service providers to a semantic interchange standard with minimal effort.
We identified the following potential partners:
- JHU APL HAIRCUT team: The HAIRCUT text retrieval system will be integrated into our semantic search agents to complement their search power in the existing non-DAML pages with our technology for searching DAML pages.
- CMU ATLAS team: Our ASCS as semantic search and translation services need to be known by other agents. We will use the advertising and matching service provided by the ATLAS to link agents and services. ATLAS will gain a translation facility that will enable heterogeneous agent to interoperate.
- Stanford DB DAML group: Ontology articulation is one of the goals of this group. We'll explore new ways to implement this core technology in our translation service.
To ensure marketability, we will do a sequence of experiments to check
(1) Better search results are achieved via our tools;
(2) The cost on performance (time, resources and security) for such results are reasonable;
(3) The use is not a burden to the end user; and
(4) The adoption of our tools is easy to the web search engine developers.
Ontology translation is essential. We can't expect all agents to speak the same language. If our translation service isn't used universally, it will only because someone else's translation service is better. The same is true for semantic search. People don't keep bookmarks for every page they want to visit. Search is an essential service on the web. With intelligent agents browsing the web in addition to humans, the issue will only become more important. There's no question that search will be needed, only whose technology solution will be employed.
5. The Next Logical Step in The 1-2 years after
We expect the semantic web and the DAML community will grow rapidly in the next few years as the demand for web knowledge sharing and semantic interoperability in various fields increases. Widespread availability of DAML annotated ontologies will emerge on the web for various businesses and their usage by the agents and applications will be ubiquitous. However, to translate the numerous ontologies and enable any two agents/applications in different ontologies communicate to each other remains a challenge in terms of time and efforts. We'll explore how to automate or semi-automate this process in the next few years.