Chapter 6 A Prototype System for Distributive Regulation Compliance Checking

6.1 Introduction

To demonstrate the methodology developed in this thesis, we implemented an Internet-

enabled prototype information management system to facilitate hazardous waste

compliance checking. The prototype system is developed using a mediation model for

organizing information sources and a distributed collaborative framework for

coordinating information flow among the waste generators, TSDS, and regulators. We

will focus the discussion on the underlying information system architecture. Examples

are provided to demonstrate hazardous waste compliance checking for waste generators.

The information for the fictitious waste generator and its waste stream data is derived

from the information obtained from the interviews in the hazardous waste TSDF industry.

The reason for using fictitious data for the research is due to the fact that all the waste

compliance checking data contain private information about waste generators and

TSDFs, and they are not supposed to be revealed to research without the permission from

the owners.

6.2 Information System Architecture

The information flow for the hazardous waste regulation compliance checking prototype

system is depicted in Figure 6.1. The regulation code is down loaded from a U.S.

government regulation repository site, in this case, from an official repository site

http://www.access.gpo.gov/ for the 40 CFR regulation code files. To illustrate the

distributed framework, we employ two computer servers, one for implementing the

1

necessary regulation code information processing and organization and the other for

compliance assistance designed for waste generators. The first server is employed to store

the original 40 CFR part 261 and part 262 regulation codes and to convert the plain text

file codes to an XML encoded semi-structured representation format as discussed in

Section 2.2. The second server is used to translate the XML based CFR regulations into

the corresponding regulation code objects, which are then stored in Oracle database for

further regulation compliance checking and information retrieval. The procedure for

organizing the code objects is implemented using the mediation approach discussed in

Section 5.2. The second server also implements the business logic for a checking service

typically employed for hazardous waste generators.

The first server, i.e., Server 1 as illustrated in Figure 6.1, provides three major

functionalities: (1) retrieving the original hazardous waste regulations from one of the

U.S. government regulation repository sites, in this case, the official web site

http://www.access.gpo.gov/ as the information source for the 40 CFR files, (2)

preprocessing the original file for the CFR regulation codes, and (3) transferring the

preprocessed XML formatted regulation code documents to client Server 2 for

compliance checking purpose. The underlying transferring mechanism used in this

system is a HyperText Transfer Protocol (HTTP) based client-server data transport

model. The procedure for the information processing on server 1 is shown in Figure 6.2.

The second server, i.e., Server 2 as illustrated in Figure 6.1, implements a number of

facilities to provide the functionalities for hazardous waste compliance checking. The

functionalities implemented include: (1) storing the preprocessed XML encoded

regulation code provision documents and translating them into database SQL language.

Using the Data Manipulation Language (DML) of Oracle database, and then importing

them to the Oracle database tables as described in Section 2.4, (2) building regulation

code objects from the Oracle database tables, (3) generating compliance checking

procedures from the regulation code objects, (4) retrieving waste lists from waste

generators, (5) normalizing particular waste lists to a canonical form suitable for

compliance checking, (6) providing compliance checking for the input waste lists, (7)

2

returning the compliance result to the client user, who is either a waste generator or a

TSDF. The information flow for compliance checking implemented in Server 2 is

illustrated in Figure 6.3.

Figure 6.1: System architecture for the prototype system

3

Server for pre processing regulation codes

Server 1

Server for providing Hazardous waste

regulation compliance Server 2

Internet

Internet

Web based compliance checking service for generators

Internet

Original regulation codes

4

Process for retrievingregulation codes

Server 1

Process for convertingoriginal file to formal XML

tagged regulation codes

Process for sending out formal XML tagged regulation codes

To server 2

Server 2

Internet connection

Inter-process

communications

Figure 6.2: Process and information flow on Server 1 for the prototype system

5

Server 1

Process for accepting theXML tagged

regulation codes

Process for converting XMLtagged codes to SQL

statement

Oracle DB tables forregulation codes

Process for generating compliance checking

procedures based on codes

Process for providing web based Compliance checking

information to generators

Process for accepting waste lists from waste generator

Process for converting waste lists to XML base

Canonical form

Process for checking waste Compliance based on checking procedures

Server 2

Figure 6.3: Process and information flow on Server 2 for the prototype system

6.3 Implementation and Features

Current implementation for the prototype system uses the latest and proven technologies:

The prototype takes advantage of the ubiquity of the World Wide Web, and adopts

the web browser as the user interface.

The modules are implemented using object-oriented programming languages.

Extendable Markup Language, XML, is used to represent and organize compliance

information.

Mediation based information model is employed for information integration.

The prototype is implemented using Java and SQL. Java is selected as the programming

language for its rich object-oriented feature and cross-server-platform support. SQL is a

standard database query language that can be employed for efficiently manipulating

regulation code provisions stored in the Oracle database. The processes shown in Figures

6.2 and 6.3 are implemented in Java. SQL statements are embedded in Java programs

through Java Database Connectivity Application Programming Interface (JDBC API) to

retrieve regulation code objects stored in the database tables.

The prototype implementation is a three-tier architecture. The first tier is the user

interface for the waste generators; the second tier is to implement the business logic and

the compliance checking related functionalities; and the third tier is for the regulation

code provisions stored in database.

The user interface, the web-based interactive checking environment, as shown in

Figure 6.4, is implemented using Java servelet technology. Java servlet provides a

component-based, platform-independent method for building web-based applications,

without the performance limitations of other web technology such as Common Gateway

Interface (CGI) programs. And unlike other proprietary server extension mechanisms

(such as the Netscape Server API or Apache modules), the servlet technology is server-

and platform-independent, and thus greatly reduces the software development,

deployment, and maintenance efforts. The servlet technology is an ideal tool for building

the prototype compliance checking system.

6

The business logics, which includes the procedures for processing the waste stream

data from the waste generators, for the information retrieval from database, and for

conducting rule-based decision making for compliance checking, are all implemented as

Java classes. Java classes are contained and run within an application server. In this

research, the Weblogic application server is used for running the Java servlets and the

related Java classes.

There are three basic modules for the compliance checking environment. Each of

them is implemented as a dedicated Java servlet. The first module, which is illustrated on

the upper frame of the web interface shown in Figure 6.4, is used for obtaining the waste

stream data from the user. The second module, which is shown on the lower right part of

the frame, is used for information retrieval for regulation codes and related information.

The third module, which is shown on the lower left part of the frame, is used for the

interactive compliance checking process. The modules are described in details in the

following sections.

7

Figure 6.4: Web browser based user interface for hazard waste compliance checking information management system

6.3.1 Module for retrieval of waste stream data from Generators

This module is implemented based on a Java servlet component model. The servlet

consists the following basic components:

The component for dealing with user input, such as (1) the location of the waste

stream file, and (2) the related regulation code the generator wish to compliance with.

The component for down loading the waste stream file from the generator’s machine

to the compliance checking server.

8

The component for converting the proprietary waste stream data into a canonical form

that can be processed by the compliance checking engine. The data conversion

process is based on a one-to-one mapping from the generator’s waste stream file to a

standard XML file. Each generator is assigned an identification number, which is

used as a key for storing and retrieving the information from the waste generator.

When a generator submits a waste compliance checking request, the identifier for this

particular generator is first retrieved and the identifier is checked against the

generator’s identifier stored in the system. Once the identifier is matched, the

proprietary waste stream data are converted to the canonical form.

The component for communicating with the module of compliance checking engine.

This component will send the waste stream data for compliance checking.

6.3.2 Module for information retrieval for a regulation code

This module is implemented using Java servlet component model for the user

interface, and using Java Database Connection (JDBC) technology for connecting to the

database that stores the regulation code objects.

The Java servlet produces standard HTML files and can be viewed by the users

using a web-browser environment. The lower right part in Figure 6.4 illustrates the

implemented regulation code information retrieval interface. This module serves as an

information retrieval help desk for the users. In order to facilitate the retrieval of complex

information embedded in regulation code, we provide multiple information retrieval

functionalities. The functionalities are classified into six categories:

Retrieval of the definitions defined in a regulation code. As discussed in Section 2.3,

the definitions within a regulation code are used to resolve the vagueness for certain

terms in the regulation code. The user can use the definition retrieval functionality of

this system to obtain the concepts for interpreting the provisions when conducting the

compliance. Figure 6.5 illustrates a compliance user retrieval request for all the

definitions that have the word “EPA” in them. The result, which is depicted in Figure

6.6, shows all the definitions that contain the term “EPA” in the definition sections in

9

40 CFR, and the explanations for the term or the phrase. The retrieval mechanism

used for obtaining the definitions from the regulation code objects is the Oracle

feature of database pattern-matching. Basically, this technique is able to search

through the regulation code objects that are stored as rows in a regulation definition

database table. The technique has been described in earlier Section 2.4.1.1.

Figure 6.5: User input for retrieval of all the definitions that have the word “EPA” in regulation code 40 CFR

10

Figure 6.6: Result for retrieval of all the definitions that have the word “EPA” in regulation code 40 CFR

Retrieval of regulation provisions by index. Plain text formatted regulation codes are

generally organized by index. As discussed in section 2.1.2, the regulation code index

is the explicit organization of the code. By providing the retrieval facility based on

index, the user can quickly find related provisions. Figure 6.7 illustrates the result of

the query for all the provisions within the “40 CFR 262.11” sections.

11

Figure 6.7: Result for retrieval of all the provisions that are in section 40 CFR 262.11 in regulation code 40 CFR

Retrieval of related provisions. Regulation code provisions contain complex linkage

structures for referencing embedded in the provisions themselves. To obtain a

comprehensive understanding of a regulation code for a certain compliance checking,

the user has to search back and forth to follow the complex references embedded in

the provisions. To simplify this type of information retrieval, we provide a retrieval

mechanism for obtaining all the linked provisions started from a particular indexed

provision. Figure 6.8 illustrates the web interface for submitting a query for obtaining

all the linked provisions starting from 40 CFR 262.11. Figure 6.9 shows the result of

the retrieval. All the linked provisions are retrieved and presented in a sequence

following the embedded references in a regulation code. For this example, the

12

embedded link 40 CFR 261.2 inside provision 40 CFR 262.11(0) is followed after the

provision 40 CFR262.11. The procedure for traversal of all the linkages embedded in

provisions has been described in Section 2.4.2.

Figure 6.8: User input for retrieval of all the linked provisions that start from section 40 CFR 262.11 in regulation code 40 CFR

13

Figure 6.9: Result for retrieval of all the linked provisions that start from section 40 CFR 262.11 in regulation code 40 CFR

Retrieval of typical normative provisions. As discussed in section 2.3, normative

provisions are the provisions that give detailed description about the compliance and

the related rules. Most provisions inside a regulation code are normative provisions.

This retrieval method allows a user to find all the normal provisions that contain a

certain word or words. Figure 6.10 illustrates the web interface for user input, and

Figure 6.11 shows the result of the normative provisions that contain the phrase

14

“hazardous waste” in 40 CFR 261 and 40 CFR 262. The procedure that establishes

meta-information for provision classification has been described in section 2.3 and

section 2.4 for the retrieval of normative provisions.

Figure 6.10: User input for retrieval of all the normal provisions that contain “hazardous waste” in section 40 CFR 261 and 40 CFR 262

15

Figure 6.11: Result for retrieval of all the normal provisions that contain “hazardous waste” in section 40 CFR 261 and 40 CFR 262

Retrieval of exceptional provisions. The exceptional provisions contain exception

rules to the typical normative provisions. To retrieve the exceptional provisions in a

regulation code is important for users to gain better understanding about the

exceptional rules or scenarios for a certain compliance request. Figure 6.12 illustrates

the web interface for user query for the exceptional provisions that contain the phrase

16

“hazardous waste” in 40 CFR 261 and 40 CFR 262. Figure 6.13 shows the result for

this query, which includes all the provisions that contain exception rules to the typical

normative provisions described in section 2.3. The classification rules and the meta-

information for provision classification have been described in section 2.3 and section

2.4 for retrieval of exceptional provisions.

Figure 6.12: User input for retrieval of all the exceptional provisions that contain “hazardous waste” in section 40 CFR 261 and 40 CFR 262

17

Figure 6.13: Result for retrieval of all the exceptional provisions that contain “hazardous waste” in section 40 CFR 261 and 40 CFR 262

Retrieval of both typical normative and exceptional provisions. The result of this

retrieval is will include both the typical normative provisions and exceptional

provisions. This retrieval function allows a user to query all the provisions that

contain a certain word or words. This function is similar to a full text search for a

word or words within a regulation code.

18

Information retrieval for a regulation code is an important part of hazardous waste

regulation compliance checking. It provides the users a tool to obtain related regulation

codes that they need to comprehend during a compliance checking process. The above

retrieval functionalities implemented in the prototype system can provide users the

flexibilities for obtaining regulation code provisions with different retrieval requests.

6.3.3 Module for compliance checking

This module is implemented using a Java component model. The Java component

model used for this compliance checking module is based on the “model, view, control”

(MVC) approach, which is a standard software design for handling user interactive input

and interaction with the underlying business logics [Alur, D, Crupi, J. and Malks, D,

2001]. For this prototype system, the approach is employed to present a user interface

(the “view”) for receiving input, a compliance checking procedure (the “model”) for

traversing a relevant rule for a related compliance code provision, and select the next user

interface (the “control”) for the checking rules based on a user’s response to the current

rule.

We implement the user interface using a Java servlet, and the compliance

checking procedure by a decision-tree based rule traversal strategy to find the related

regulation provisions. The logics are implemented in Java classes. The Java servlet acts

as an information controller for creating user interface layout, for retrieving user input,

and for invoking the backend regulation code traversal component. More specifically, the

servlet (1) receives user compliance checking requests from the client, (2) selects the

proper regulation code to use for compliance checking, (3) performs decision-tree based

traversal through the regulation compliance rules for clients based on the procedures

developed in section 3.3, and (4) selects and delivers relevant rules for regulation code

provisions to the clients for further actions. These tasks are performed iteratively until all

the related regulation rules are traversed and checked. In Figure 6.14, the lower left part

19

illustrates the user interface of the compliance checking process module. The user

interface is composed of two parts. The first part is the display of the retrieved regulation

code together with the explanation of the provision. The second part is the decision

choices for displaying the provision. There are three decision choices, namely an answer

“yes”, an answer “no”, and an answer “unknown” for the displayed provision. The Java

servlet will retrieve the choice from the user and perform further rule traversal for a

particular compliance checking.

Figure 6.14: User interface for conducting compliance checking rules for waste generators

Following the above example, the compliance checking module will present a

user the checking sequence for determining if the waste lists submitted by the user is a

hazardous waste based on the interactive input from the user. The complete checking

process is depicted in Figure 6.15.

20

Figure 6.15: Sequence of checking rules for determining a hazardous waste for a generator

6.4 Summary

This chapter provides a description of the prototype developed in this research for an

Internet based information management system for hazardous waste regulation

compliance checking. The focus of the prototype system is to demonstrate the feasibility

of using the methodology developed in this research for the implementation of the

information management system. In addition, we use the prototype system to obtain

insight and experience for developing software components for distributive information

systems using current state of practice.

In the current implementation, information retrieval for regulation codes and the

related information organization and management for code objects is emphasized. The

rule-based checking procedure, however, is implemented using a simple decision-tree

based traversal to obtain the relevant rules for a certain compliance checking, but lacks a

21

Step 1

Step 2

Step 3

Step 4

Step 5

Step 6

Step 1 : beginning check process for the waste. Step 2: checking if the waste is excluded from being a waste.

Step 3: checking if the waste is a simple waste. Step 4: checking if the waste is a solid waste.

Step 5: checking if the waste is a listed hazardous waste.

Step 6: Concluding that it is a hazardous waste since at least one component of the waste is a hazardous waste.

formal rule-based system for conducting inference, conflict resolution for complex rule

sets for more general compliance checking cases. In future work, an investigation of

integrating a rule-based engine with the information management system for compliance

checking will be beneficial for the research in regulation code based compliance checking

system as well as the practices in the regulation compliance service.

22