Automated Text Categorization (ATC) is a useful technology to build such software tool that can categorize a document to one of many predefined categories. Unfortunately there is no as such classifier for Holy Quran, one of the most important documents of the universe. This is because Holy Quran is written in Arabic and Arabic language processing is yet not that mature that language processing can be done. This paper aims on building a software tool that can categorize any verse of Holy Quran to one of predefined categories and if it is not falling in one of predefined categories it can automatically define a new category. New category will be added to database for future categorization process. Moreover, this categorization is not just based upon word count rather it is based upon word as well as meaning of the words in the verse and there correspondence in semantic network. The semantic network for this task is created and used in this classification. This tool will help those people who want to know the theme of a verse of Holy Quran.

A learning apprentice system is presented, which leams from examples extracted from user dialogues. The system provides an interface between the user and the turbine modeller. While a dialogue is carried out between the user and the turbine modelling software, the system observes the dialogues and whenever a new example is observed which performs a task completely, the system tries to learn it. The learning methodology used by the system is described and various drawbacks are pointed out. A new learning methodology is proposed which easily overcomes the problems faced by the earlier methodology.

This paper describes successful implementation of a Rule Based System at MTBC, for applying billing compliance rules on medical claims. Rule engine has been developed in Structured Query Language as stored procedures, which is one of the unique features of this rule based system. Implementing rule engine in SQL has provided two major benefits. Firstly, as operational data of the organization is in relational form, stored in Microsoft SQL Server database, therefore rule engine, using the native language, works at real time without any need of data transformation for working memory. Secondly due to SQL server, rule engine is using interpreted approach instead of compiled approach, which helps dynamic updating, editing and execution of rules. A rule is represented as a query stored in database, along with associated attributes like rule name, rule description and rule priority. 'where' clause of query contains condition and then part of the rule. A rule editor has been developed to facilitate domain users to edit rules in English like format, which is then translated to SQL statements. Editing of business logic has become very easy in MTBC billing software by using MTBC-RBS.

No abstract