Kyle F. Morton

The ability to store XML in a database for retrieval and updating on demand has became a requirement. Also there is the ability to take the inverse of that requirement which is to take the regular or nonXML data and convert it to XML form. There are three ways that data can be stored in a database which are:

1. Store the entire document as attribute values in a table.
This is done by defining a new data type “XMLDOC” which values are XML documents and define specific attributes of specific relvars to be of that type. XMLDOC have a set of associated operators that provide retrieval and update capabilities on XMLDOC-valued attributes at a lower-level, supporting access to individual elements or XML attributes.

2. Shred the document and represent various pieces of it as various attribute values within various tuples within various relations.

Although no new data types are used XML documents are shredded into individual XML elements and attributes and stored as values of various relational attributes in various places in the database. With this approach an application can create an XML document from regular database data. The problem know of method is the “impedance mismatch”. This is when child nodes of a given parent node in the hierarchic model form not a set but a XML document D, where aspects of D is lost when shredded and stored in the database.

3. Store the document in a native XML database and not in a conventional database.
There is no reason to use a XML database which is just a desire for completeness. A relational model is adequate enough to represent any data.

I have redirected my research temporally to educational based search engines. My focus is to develop a search engine that will promote learning. Currently, there are downfalls with most modern day search engines, which are:

1. vague array of information
2. doesn’t use reasoning to consolidate information to formulate a search result
3. search results are not narrow or specific
4. results may not be relevant to keywords or phrase
5. does not show mathematical steps of computed results
6. hinders the student’s ability to compute equations mentally

The key is to give a user relevant results related to their search topic. You don’t want to give the user over 15 search results because it will then become a tedious process of elimination for the user. But, you do what the user to go through a thinking process to find exactly what they want. If the search engine was to just give an user an answer it will create negative impact on users learning ability. For example: most people have lost the skill to computed mathematical equations mentally because they have been crippled by calculators.

My current research will apply solutions to address these problems while  increasing search engine accuracy.