Better Database Designs Quickly And Easily From Simple English Sentences


by Venkat Reddy - Date: 2007-02-28 - Word Count: 718 Share This!

1. Introduction and overview

I first present four common problems with the existing database design methods. To overcome these problems, next I introduce you a new method of building directly database designs from simple simple sentences. Also, I provide the benefits of the new method to your software development team.

2. The 4 common problems with database design methods

Database design methods, such as ER modeling, etc., have following 4 common problems in directly building designs from software requirement specifications. Make sure you have spotted them too.

The methods are ad-hoc and require longer learning time. This is because these methods have no clear rules to decide entities and their attributes from nouns that occur in software requirement specifications.

Messy entities cause nightmares to maintain applications. Such designs are possible since there is freedom to introduce foreign keys in entities.

Poor communication across all stake holders of a software project. This is because database design is an abstraction of information with a focus to represent data. So it represents only partial information which can be misinterpreted. A worse example of a fatal error is the explosion of ARIANE 5 due to misinterpretation of data as signed or unsigned ones.

Validating and maintaining database designs are slow against software requirements and their changes.

A new fact oriented analysis method that builds database designs directly from simple sentences is the answer.

3. A quick overview of the new method

The new fact oriented analysis method has three main steps: 1. Writing facts from an understanding of software requirement specifications, 2. Analyzing facts, and 3. Generating a data model from analyzed facts.

For your quick overview, I give analyzed facts and a UML data model generated using the new method. The facts pertain to the problem domain of an office.

From looking at the facts, you recognize that the first challenge is to write single clause sentences. You can see that writing them is really easy. But challenge is in recognizing them. Even if you do a mistake in identifying and stating them, it is also easy to review and correct them.

Next, you can notice that some text in these facts is in bold font face. This is all about visually indicating nouns. Also you can notice that the bold font text comes with additional font styles such as italics, and underline. These additional styles are to visually distinguish noun classifications.

Finally you can also notice that each of these facts have a graphic icon on its left side. These icons are to indicate the facts' classification.

In summary, your job as analyst, database designer or information architect is to find and write all the useful facts; analyze them to find their nouns and their classifications; and also classification of facts.

When all the facts are analyzed, a tool like EnglishToUML can be ready to generate the data model using the new fact oriented analysis method.

I have chosen to provide UML data model instead of a ER model because of a tool support for the method. Both these models are used to represent database designs.

You can view these ideas in action in five minute demo (Windows only) of EnglishToUML.

4. The 3 more benefits you can adopt for your team

We see the following 4 benefits from the new method and its tool for data modeling.

Enhanced database Modeling productivity. Experienced data modelers efficiency to go by 20% - 30%, and for beginners by 30% - 50%. We shall be able to convert requirement specifications to data models with less experienced engineers. Also expect 70% to 80% improvement in teams collaborative effort as team members can independently collect facts, and later collate them using a tool like EnglishToUML.

Savings of time. Productivity implies savings of time too. This mean business analysts can save 20% to 40% because of improved communication with customers and domain analysts, and the rest of team members such as software architects, software engineers, and maintenance engineers to a tune of 20% to 50%.

Faster and cheaper to train new data modelers. This is mainly because we have turned around analysis problem as natural language exercise of writing single clause sentences. 5. Conclusions

I have briefly introduced a new method of directly building good database designs from simple English sentences. The method is easier to learn and practice. And it improves your teams productivity by at least 20% while overcoming the common problems in designing good databases.


Related Tags: design, data, english, analysis, database, models, uml, english to uml, database from english

About The Author:

Dr. Venkat Reddy, CEO, ObjectTrees Soft Consulting, is the inventor of the new analysis method and the coder of its EnglishToUML tool. You can reach him at http://www.EnglishToUML.com.

Your Article Search Directory : Find in Articles

© The article above is copyrighted by it's author. You're allowed to distribute this work according to the Creative Commons Attribution-NoDerivs license.
 

Recent articles in this category:



Most viewed articles in this category: