System: web enabled agent-based university timetabling assistant
Contents
1. 58 PROJECTPLAN 58 PROJECT RESULTS AND PRODUCTS Dt demo t etum st 59 RESOURCES REQUIRED os quiu auff boot aut 59 a xd tiles 60 OSV 60 APPENDIX II EET 61 SOURCE cov iod caftro p S SO 7 61 User caseus adde 67 PRE Scd mE 13 LIST OF ABBREVIATIONS NP MAS CHC HC SC AMAS NCS Non Deterministic Polynomial Multi Agent System Common Hard Constraints Hard Constraints Soft Constraints Adaptive M ulti Agent Systems Non Cooperative Situations CHAPTER 1 INTRODUCTION 1 1 BACKGROUND 1 1 1 General Timetabling According to Collins Concise Dictionary 4th Edition a timetable is a table of events arranged according to the time when they take place The events are usually meetings between people at a particular location Consequently a timetable specifies which people meet at which location and at what time A timetable must meet a number of requirements and should satisfy the desires of all people involved si2. Thestudents groups A student group needs to take up a certain number of courses in a particular semester whereby each course is allocated a number of time slots in the timetable to complete the syllabus X time slots for a course 1 Y time slots for a course 2 etc lil lecture rooms Lecturer rooms can either be equipped or not equipped The lecture rooms that are not equipped are normally used to carry out the theoretical part of a course The lecture rooms that are equipped have tools necessary to undertake a practical and these rooms can only be available at a given time slot or on a certain day because some of these facilities are shared among different student groups The success of this system will depend on how well the system prioritizes and executes the constraints 4 1 2 User requirements The system will be deployed in two schools within the university namely the School of Computing and Informatics and School of M athematics The following will be the users of the system e The lecturers The student groups e Course manger e Rooms manager 4 1 2 1 Functional Requirements The following are the functional user requirements v The lecturers will need to declare their time preferences and also state their back up options in case that particular time slot is already taken v The students from various student groups will need to view the end result timetable v course manager will need to allocate different
3. research done for the project And last but not least would like to show great appreciation to my family for being there at all times and for the unconditional support they have given me throughout this period Thank you all Stephanie Gaku ABSRACT The university course timetabling problem is a difficult problem that most universities face on aregular basis It is considered as an iterative and time consuming process for it involves a lot of negotiating and compromising of resources in order to resolve any conflicts that may arise during the process Currently the generation of these timetables is done manually and the results are often not satisfactory as they do not fully optimize the institutions resources Some educational institutes have tried solving this problem by the automation of this process whereby mathematical models are used to deal with the scheduling problems Such an approach requires the generation of multiple mathematical models so as to satisfy the requirements of different educational institutes Therefore a web based multi agent timetabling assistant is proposed to solve the timetabling problem In this case an agent is a computer system that is situated in some environment and 15 capable of independent action so as to meet its design objectives The proposed system will use agents to compete for a number of resources on a given number of time slots whereby negotiation will be used as a means for reaching comprom
4. 24 CHAPTER 3 METHODOLOGY 3 1 PLANNING AND SELECTION The feasibility analysis carried out for the project is as discussed below Economic feasibility In the initial stages the system was economically viable to create for it only requires the programming tools Once the system was created will be taken out for testing Technical feasibility The system was deployed using a web platform which makes it easy to access and readily available to the public to use Operational feasibility it will require adequate skills in web browsing so as to experience the efficiency and the effectiveness of the system Scope The scope of the system will extend to data collection analysis and representation of data the website The system for now will focus on two schools i e School of Computing and Informatics and School of athematics with time the system would be Inco operated to the university as a whole 3 2 DATA COLLECTION The primary means of data collection that was used so as to acquire the users needs was the use of questionnaires and interviews The interviews that were carried out were used to acquire the techniques used to come up with a timetable manually The logic behind coming up with the timetable manually would then be expressed into program logic and later be used to automate the whole timetabling system This information was then generalized to come up with an online system that would be able to addre
5. Furthermore each agent can adopt a different algorithm to create a hybrid and parallel course timetabling solution model In such an operation mode the system control would be much more complex but it could be expected that the problem may be solved more quickly and more effectively 2 5 THE AMAS ADAPTIVE MULTI AGENT SYSTEMS THEORY The AMAS theory provides a solution to build complex systems for which classical algorithmic solutions cannot be applied The concerned systems are open and complex All the interactions the system may have with its environment cannot be exhaustively enumerated also unpredictable interactions can occur during the system functioning and the system must adapt itself to these unpredictable events The solution provided by the AM AS theory is then to rid ourselves of the global searched goal by building artificial systems for which the observed collective activity is not described in any agent composing it i e Each agent only pursues an individual objective and interacts with agents it knows by respecting cooperative techniques which lead to avoid unpredictable situations like conflict concurrency etc called Non Cooperative Situations 18 NCS Faced with a NCS cooperative agent acts to reach a new cooperative state and permanently adapts itself to unpredictable situations while learning on others Interactions between agents depend on their local view and on their ability to cooperate with each other Ch
6. Previous timetables that were manually schedules 59 BUDGET ITEM QUANTITY PC LAPTOP 1 Patch cable Flash disk Printer cartridge Rim of paper 1 Software Ms office Students Whole suite version Antivirus Travel cost Within Nairobi Communication Local calls Training Miscellaneous CONCLUSION PRICE in KSHS 35 000 500 1 500 3 000 525 1 000 10 of total TOTAL 35 000 00 1 500 00 6 000 00 525 00 1 000 00 1 200 00 4 572 50 50 297 50 This project will be able to demonstrate how the course timetabling problem at the university can be modeled as a constraint satisfaction problem The use of agents suggests that the system is highly flexible and can be easily maintained due to the highly adaptive nature of multi agent systems M AS These qualities prove to be which Is crucial for any university 60 APPENDIX III Data Source Sample Log In Execution Code lt php Start session session start Include database connection details require once config php Array to store validation errors errmsg arr 11 Validation error flag false Connect to mysql server link mysql_connect DB HOST DB USER DB PASSWORD 4 die Failed to connect to server mysql error Select database db 2mysql select db DB DATABASE 61 if db 4 die Unable to select database Function to sanitize values received fro
7. UNIVERSITY OF NAIROBI SCHOOL OF COM PUTING AND INFORMATICS SYSTEM WEB ENABLED AGENT BASED UNIVERSITY TIM ETABLING ASSISTANT BY STEPHANIE GAKU This project has been submitted as part of the fulfilment of the degree of Bachelor of Science of the University of Nairobi JUNE 2012 DECLARATION The project presented in this report is my original work and has been done with full support of my supervisor as part of the fulfilment for Bachelor of Science in Computer Science Gaku Stephanie P15 23197 2008 This project has been submitted as part of the fulfilment for the Bachelor of Science in Computer Science with my approval as a lecturer in the University of Nairobi School of Computing and Informatics and as the Project Supervisor Mr Elisha Opiyo School of Computing and Informatics University of Nairobi ACKNOWLEDGEMENTS First and foremost would like to thank God for giving me the wisdom and strength to develop this project to completion also want to thank my supervisor Mr Elisha Opiyo for being available and resourceful at all times in the development phases of this project His support and ideas were major driving factor for this project also thank my fellow classmates for their support and conducive environment they offered for the period undertook this project also thank School of Computing and Informatics for convenience in opening the library and the computer laboratory which were imperative in
8. we Fi _ emm oe Eo Stephanie SW100 Negotiate Time Slots List of Slots you d like to negotiate 1 Slot Availability nen DENT ncn om ferfier SS jenen _Non negotiable 69 View Schedule SCHOOL COMPUTING AND INFORMATICS Cc Stephanie My schedule Course ID Course Name Vear Room 05041 Computer Network FICHRM 5 414 ICTs and Socie 4 FICHRM Rossen __ PER The administrator login allows the administrator to allocate rooms and edit other system information Allocation of rooms step 1 SS 4 E Y S d ABOUT US ARTICLES CONTACTUS SITE MAP l gt Rooms Welcome SOM 4 admin Degree Courses b Allocate Rooms Log Out uen Room Allocaction Step1 Student Group Year 1 v 70 Allocation of rooms step 2 Welcome SOM 5 Room Allocaction Step 2 Allocate Rooms gt Log Out Student Group Capacity Number of Students of Stra Please Select Room RoomID Finish Back to Previous Page The users are able to navigate through these pages and perform t
9. Informatics and the School of Mathematics The information collected will then be used to form constraints for the system so as to make the system as practical as possible The system will have the following modules The course booking module this module will enable lecturers to book whatever course they d like to teach The time slot booking module this module will enable the lecturers to book time slots of their preference depending on the availability of a particular time slot The negotiating module this module enables lecturers to negotiate time slots in case they are not satisfied with the current time slot they possess The timetable generation module this module is able to generate the class timetable the course timetable and the lecturers own schedule The administrator module this module enables an administrator to add edit delete and search for various components of the system This module also supports room allocation to various student groups The student module this module enables a student to register for courses and also view their respective timetables The visualization tools this module helps visualize the availability of timeslots and also their negotiating state 12 CHAPTER 2 LITERATURE REVIEW 2 1 INTRODUCTION Generating a timetable from scratch consists of scheduling a set of courses within a given number of rooms and time slots and due to the wide number of constraints the generic solutions for suc
10. consecutively There is only one CHC Agent in the system The number of HC and SC Agents depends on the specific course timetabling problem The constraint agents are independent of one another The system design is presented in three different modes of operation the passive mode the active mode and the parallel active mode 15 2 4 1 Passive Mode Notepad Course timetable lt gt Communication between agents The passive mode general solution model is composed of one CHC Agent one HC Agent and a number of SC Agents The hard constraints can be represented by 1 HC agent or a couple just as the Soft constraints SC but in order to simplify the illustration 1 is used Generally there are more than one soft constraint in the course timetabling problem so the soft constraint agents are represented by the SC 1 Agent through SC n Agent where the value of n depends on the number of soft constraints During the course timetabling process the CHC Agent first produces a basic feasible course timetable consistent with the common hard constraints Thereafter it sends a copy of the basic feasible timetable to each of other constraint agents Each constraint agent evaluates the timetable based on its own constraint and sets penalty points on the number of constraint violations These evaluations are sent back to the CHC Agent and it develops an overall evaluation result on the basic feasible timetable Following this the CHC Agent creates a
11. processing speed automatically enhances Its script is optimized to make the server s job easier thus nowadays the uses of PHP 15 being popular among the programmers 5 4 2 JavaScript is a scripting language developed modify web pages In this project JavaScript was particularly used for the PC web interface to increase usability of the web based timetabling assistant system in terms of the look and feel of the web interface 5 4 3 HTML Hypertext M arkup Language HTML enables you to markup text so that it can function as hypertext on the web that is text with hyperlinks that allows for navigation All web browsers are based on HTML 5 5 Coding The coding effort involved a structured and modular approach which functions procedures needed were identified and written individually All functions were written individually and stored in separate files The following diagram shows design of how the Web module was implemented Website M ap 3 Student RocateRoons je le Register for Courses ModiyGouses View Cass Timetable Negotiate Times iege Course Vewschedue Generate reports View Timetables Overall design Architecture Request Web 5 6 Testing Testing system
12. scheduling 5 44 memes 14 2 4 AGENT SOLU ION sp 15 PES MN MO Sa otc ete or E 16 Lp Wenn E E 17 2 4 3 Parallel active 6 18 2 5 THE AM AS ADAPTIVE M ULTI AGENT SYSTEM S THEORY 0 00 18 aat DC LOGY M 19 NS DW 23 DORE IED 23 Monash University Timetable System 23 DAES OSS RET 24 24 Mimosa Scheduling 5 1 a nennen neni 24 CAP ego Q 24 CHAPTER ee Wit ODIO 25 ded Ctra dandi 25 32 DATA COLLECTION siia 25 METHODOLOGY osaa od 26 CHAPTER 4 ANALY SIS AND DESIG 28 ZEPREDEIM INARY ANALY SIS 28 d T1 E Resultsor CIS Thntel VIBW outs 28 2 1 2 User regui emeni S cte epi aq od M b ea ai e utat Cl ud bei er re 29 4 1 3 The basic concept of the system architecture 30 4 1 4 The demands imposed by the fac
13. small disturbance to the timetable and guarantees that the disturbance does not violate the common hard constraints The small disturbance can be realized by either moving one course to another available position in the timetable or exchanging the positions between two or three courses As a timetable improvement proposal the disturbance is sent to the constraint agents for evaluation The constraint 16 agents evaluate the proposal and provide feedback to the CHC Agent After developing new overall evaluation result on the modified timetable the CHC Agent decides if the modified timetable represents an improved schedule If the modified timetable is accepted the CHC Agent will adopt it as the current best timetable and inform other constraint agents to make the same change In this way the system can guarantee that at any moment the course timetable carried by each agent is exactly the same If the modified timetable is rejected the CHC Agent will discard it The CHC Agent then continues looking for a new proposal attempting to further improve the timetable This process is repeated until all the constraints violations are eliminated and a final course timetable satisfying all the constraints is found On the other hand if the constraints violation cannot be reduced further the course timetabling process will go to the active mode described in the next subsection In this timetabling mode only CHC Agent makes proposals and all the other con
14. to the College of Physical and Biological Sciences that comprises of various schools that individually deal with their own majors For every major there is a set of associated courses which can either be compulsory or elective The lecturers are shared between schools depending on the courses they teach and there are incidents whereby students from different schools need to undertake a common course that will require students to occupy shared room resources The timetable has a span of five days starting from M onday to Friday for the undergraduate students and each day is divided into various time slots whereby the time slot values varies from school to school The events that can take place in a time slot can either be a lecture or a practical session and is given by either a lecturer or a teaching assistant The lecturers take place in a classrooms or lecture halls and the practical sessions take place in the labs In summary an event is given by alecturer or an assistant during a time slot in a day to a specific group using a specific room resource This relationship is represented by a timetable for all events provided that hard constraints are not violated Hard constraints are constraints that cannot be violated and are considered to be of great necessity to the university operation These constraints can be also added by some lecturers like part timers or visiting professors The timetable tries to satisfy other constraints which are not very i
15. was carried out using regression testing This is done by testing each module independently and then integrating the modules to ensure that they work properly together and cohesively The testing was iteratively as each component was added To test the web system different values were entered to ensure that the system has been properly validated and the user cannot enter values that aren t correct Appropriate error responses were also created so that the user can know that the transaction performed on the system wasn t successful and whether they should try again later or use contact the administrator for help Debugging of the application was done so that the system does not crash when it incurs an error 49 5 6 1 Function Testing Functions and procedures were tested one at a time Their outputs were displayed to verify they were working In the intended manner Examples are shown below This message 15 displayed when a user has successfully registered to the system ap Chap TIMETABLING TIMETARI E GE OMATIC TIMETABLE G AX n ABOUT US ARTICLES CONTACT US SITE Registration Successful You have succesfully reigistered to the site click here to login This message 15 displayed when a lecturer has successfully booked courses of their preference and based on their availability ABOUT US ARTICLES CONTACT US SITE MAP Book Courses Book Timeslot
16. 9 Complete design diagram Grid 1 Co operative Agent lt lt Agent gt gt 1 Ones Day 1 Booking Agent Representative lt lt Agenb gt lt lt Agent gt gt Dim Object 1 1 Constraint Managger DimObject Configuration Manger Window Converter Course Manager Agent Time Slot Student Grooup 45 CHAPTER 5 IMPLEMNTATION AND RESULTS 5 1 Hardware Platform used e Personal computers and laptops running windows operating system Web Servers and backup servers running on Apache Processor e Intel amp Celeron amp CPU 2 19 GHz RAM 1 87GB e System Type 64 bit Operating System e 150 GB Hard Disk 5 2 Software Platform used e Adobe Dreamweaver 53 e Macromedia Fireworks Xampp Server e MySql for the database e language used JavaScript PHP has an active community support on the Web e Visio Professional Visual Paradigm e Microsoft Word 2007 e Microsoft PowerPoint 2007 Operating System Windows 7 46 5 3 Network Platform used Timetabling assistant system is a web platform and thus requires the following network configuration of facilities available a WAP EDGE 3G 3 5G for mobile devices to access the site b Any internet connection to access the PC website 5 4 Choice of programming tools Web programming technologies tools and languages were used to implement the platform The following gives a description and basis for selecting the
17. Action An action is a means to act on the environment during the action phase of an agent An attribute stereotyped with action stands for a parameter of an action For example a move length the maximum size for a message A method stereotyped with action represents a possible action for an agent For instance move send a message An agent is the only one which can use its actions they are then private Rules o an attribute which is stereotyped with action is private o action stereotyped method is private o amethod which is stereotyped using action can only be called during the action phase of an agent 20 Skill A skill stereotyped attribute represents a datum which is useful to act on the world or a parameter that is linked to a skill stereotyped method For instance an integer value represents a minimal distance a robot hasto respect to avoid obstacles A skill stereotyped method represents a reasoning that the agent can only make during its decision phase For example a method that describes the logical reasoning to avoid obstacles A skill stereotyped attribute or method can only be accessed affected or called by the agent itself to express its autonomy of decision Skills can be represented by a multi agent system when they need to evolve Rules o an attribute or a method that is stereotyped with skill is necessarily private o aSkill stereotyped attribute can only be used by a skill stereotyped m
18. Booking Agent Class2 p 1 1 Course Manager Representative Agent Class1 Represents an agent 4 3 8 Agent Architecture The BookingAgent specification Skills e Aptitudes and interaction languages are methods e Representations are attributes oom 1 1 1 Room Manager Booking Agent Representation negotiating partners currentCell e Constraints bookState negotiatingState movelnGrid manageNegotiation manageConstraints manageBooking Amplitude bookATimeslot cancelBooking negotiateTimeslotBooking CancelPartnership negotiatePartnership establishN egotiationPartnership 43 Non Cooperative Situation Model These are the different situations an agent might come across that may lead the agent to a Non Cooperative State and the corresponding actions they should take to come out of the NCS In this case the NCS for a BookingAgent are Partnership incompetence the BA meets another BA that may an uninteresting partner Booking incompetence the BA is in a cell that is uninteresting to book Partnership conflict the BA meets another BA that is interesting but this other has already partner Booking conflict the BA is in a cell that is interesting to book but this cell is already booked Booking uselessness the BA meets its partner they must separate to explore more efficiently the grid 44 4 3
19. CHAPTER deuda qu Moti 46 3 L Haraware PIAS 46 3 2 SortWare PIatForTidised a e eR 46 NSCWOPK Plat Oni etd 47 EOS saquen Mtas _______________ 47 oo e ch 47 DAZ AVASERIP 48 T 48 Es OCI rae 48 DO A 49 ws Oe Ul TESTING __ __ 50 2 Dr 52 p 935 SEG ______ _6_________ 53 5 ES ah gu erm 54 CHAPTERS DISCUSSION titres 55 eMe eai s CIN SINC ERE DOO EA 55 smod C 55 c 55 6 4 Suggestions TOF WOK ___ FU DU 56 CISION T 56 57 APPENDIX Em 58 PROJECT PLANNING AND RESOURCES
20. D WORK Monash University Timetable System Monash University Timetable System is a web based timetabling system that is able to generate timetables from scratch using different constraints such as user defined preferences The system then outputs the generated timetable with reference to the users preferences The system has been used for quite some while and has significantly improved the general performance of the schools management Below is a link to the Monash Timetabling System https mutts monash edu au mutts timetable bymodule aspx splus 2012 23 Untis Express Untis Express is a timetabling assistant that enables one to enter basic data about the school define exactly which teacher is available at what time avoid the scheduling of lessons at a particular time in is able to fact in all constraints and come up with an efficient timetable using a self engineered algorithm The algorithm of Untis Express does not calculate just one but many different results then gives you the option to choose the one which you like most by defining exactly which criteria are the most important ones and which ones can be broken necessary FET FET is open source free software for automatically scheduling the timetable of a school high school or university It uses a fast and efficient timetabling algorithm The fully automatic generation algorithm allows also semi automatic or manual allocation Usually FET is able to solve a compl
21. anging these local interactions reorganizes the system and thus changes its global behavior Applying the AM AS theory consists in enumerating according to the current problem to solve all the cooperative failures that can appear during the system functioning and then defining the actions the system must apply to come back to a cooperative state 2 6 ADELFE METHODOLOGY The following are the activities that are involved when using the ADELFE methodology to develop a multi agent system 1 Define the studied system _ 10 Express the detailed architecture 2 Determine the system context the agent model _ 3 Determine the entities 11 Give each agent architecture e 4 Characterize the environment d _ 12 Express NCS 5 Express the use cases 13 Give class diagrams Design Workflow Analysis SA Workflow E 9 Study interactions between components 9 8 Identify agents T Verify the AMAS adequacy gs Analyse the domain and identify components Workflow Designing agents In the ADELFE methodology nine stereotypes have been defined to express how an agent is formed and or how its behavior may be expressed Co operative agent A cooperative agent 15 an agent that possesses a cooperative social attitude As an object a cooperative agent must have a run method which simulates its lifecycle which consists in having perceptions taking decisions and then doi
22. ble from scratch given a number of constraints such as time number of students classrooms capacity lecturers availability etc In order to effectively achieve the requirements of the time tabling assistant the following objectives have to be met e Undertaking an understudy of the university scenario so as to come up with clear timetable requirements Designing a system that proves to be flexible and scalable with increased collaborated use of assignments e Implementing the timetabling system and ensure that its able to o Maximize on the schools human and technical resources Reduce number of conflicts that may arise when developing the timetable o Ensure fair re assignments of resources where the demand of utilities is matched to the supply Testing the functionality of the system by conducting interviews so as to get people s views and opinions about the system Report and document the users feedback 1 4 SYSTEM JUSTIFICATION The web enabled agent based university timetabling assistant will largely contribute to the generation of feasible timetables by using negotiating agents that are able to compromise conflicting interests in the shortest time possible This will in turn save weeks of manual work and effort and also improve the quality of education since the universities facilities will be fully optimized 11 1 5 5 This system will focus on the collection of data from the School of Computing and
23. courses to the lecturers depending their capability to teach that course v The room manager will need to allocate classrooms to student groups depending on the capacity of that particular student group and the nature of that course practical lesson or theoretical lesson v system should provide a timetable that is able to meet the users needs and preferences 4 1 2 2 Non functional requirements Usability The system should provide a friendly graphical interface to ensure ease of use when the end users utilize system functionality The interface to the system should be inherently fail safe full proof and overly cautious in defending against accidental and intentional misuse Efficiency The system should support a response time for addressing severe issues in less than 1 minute Availability The system should be protected against both accidental and malicious denials of service and is available for use whenever it is expected to be operational Confidentiality Only legitimate officials have access to the users confidential files and data Integrity The system should be tamperproof whereby system changes are only executed by system administrators Security Only legitimate officials have access to the timetable database and this would be achieved by giving administrative passwords and usernames 29 4 1 3 basic concept of system architecture The system is web based hence making it readily availabl
24. der location login failed php exit j lelse 1 die Query failed j j if mode lecturer Create query qry SELECT FROM lectusers WHERE username username AND password md5 POST password result2mysql query qry if result 4 if mysql_ num_rows result 1 Login Successful session regenerate 100 member mysql_fetch_assoc result 64 SESSION SESS M EM 101 2 member lectmemlID SESSION SESS FIRST 2 member fn SESSION lectid 2 memberr lectid session write close header location member index2 php exit lelse Login failed header location login failed php exit j lelse 1 die Query failed j j 5 admin Create query qry SELECT FROM admin WHERE username username AND password md5 POST password result2mysql query qry if result 4 if mysql_ num_rows result 1 Login Successful session regenerate 100 member mysql_fetch_assoc result 65 SESSION SESS M EM ID 2 member adminld SESSION SESS FIRST NAM E member username session write close header location member index3 php exit lelse Login failed header location login failed php exit j lelse 1 die Query failed j gt 66 User Manual This section gives a description of the procedures to be followed when using the Chap Chap Timetabl
25. e environment The environment in which the system runs 15 said to be e Dynamic the evolution of the active entities does not depend on the system they are unpredictable from the point of view of the system e Accessible the environment can obtain information on the state of the environment Non deterministic the system is not able to know what could be the effects of its actions on active entities e Continuous the number of interactions between system and entities are infinite 4 2 2 System entities There are two types of entities within the system a Active entities they are entities that are able to change by themselves change their own constraints or they can interact with the system In this case they would be lecturers student groups the course manager and the room manager b Passive entities they are entities that represent resources and their character cannot change on their own In this case it would be rooms and the database that contain information about courses number of sessions per week duration of each time slots number of courses earliest session latest session free periods etc 4 2 3Context definition The collaboration diagram between the passive entities and the system precondition Lecturers capabilty to teach a course 1 Request to teach a course 2 Check if the course is available gt 6 Allocate course to lecturer 3 Return result Lecturer Timetabling sys
26. e physical environment or the agent itself These representations are used by the agent to determine its behavior An attribute which is stereotyped with representation is a means to express a knowledge unit that describes an agent For example the number of agents known by an agent is knowledge A method which 15 stereotyped with representation corresponds to a handling of a representation access alteration For example a method which allows alteration of a user profile An attribute or a method which Is stereotyped with representation can only be accessed affected by the agent Itself Rules o An attribute or a method which is stereotyped with representation is necessarily private Arepresentation stereotyped attribute can only be used by a method which is stereotyped with representation or aptitude o Arepresentation stereotyped method can only be called during the decision phase of the agent Interaction The interaction language represents tools that enable an agent to communicate directly or not with others or with its environment A method stereotyped with interaction expresses the ability an agent has to interact with other agents or with its environment in an indirect manner An attribute stereotyped with interaction represents a functioning datum or a parameter of an interaction made via an interaction stereotyped method Rule o Amethod stereotyped with interaction can only call methods ste
27. e to its end users via the internet Each lecturer will need access to the system using unique usernames and passwords Thereafter they will be assigned courses to teach using the course manager depending on their capability The lecturers will then be able to choose a time slot that they prefer teaching based on their availability and this would be done through the help of the bookie The room manager will then need to allocate rooms to student groups based on the capacity of the student group This information will then be stored in the systems database and then used to come up witha suitable timetable that satisfies the users needs and preferences Once a suitable timetable is generated both the lecturers and students need to view it using their unique usernames and passwords to gain system access Request course Course Manager Check capability 1 Time slot prefrence View Bookie Book Time Slot Cell Student Capacity View Student Check availabiliy Rooms Manager Represents an agent 4 1 4 The demands imposed by the facilities and human resources There are a number if constraints that need to be addressed in the development for this system e The lecturers availability in terms of time e The lectures capability to teach a certain course The rooms capacity The rooms state equipped not equipped 30 4 2 SYSTEM REQUIREMENTS 4 2 1 Characterization of th
28. equest granted if no of students gt room capacity split student group Return student capacity Room big enough consider splitting student group 35 4 2 4 Determination of use cases Initialize courses Modify courses N Initialize N constraints N Top Package Lecturer Launch solving Top Package Course manager 7 7 7 7 7 7 Vizualize current result t9 m T 2 em 7 m E Co operation m d Failure g Possibilities Initialize rooms Cuan Top Package Room manager Modify rooms Top Package Student 36 4 2 5 Identification of co operation failures The associations to Visualize current result case are potentially non cooperative the result of the time tabling resolution is the only cause of cooperation failure between the users and the system in the sense that users expect the system to satisfy their constraints Here are some of the situations that may result to non co operation The lecturer is not able to get a time slot that they desire The lecturer not able to teach a course that they desire Some courses not getting lecturers to teach them cause the remaining lectures are not qualified to teach them and the ones capable have taken
29. er on what to do Below is an illustration of this function ABOUT US ARTICLES CONTACT US SITE MAP B x u 2 gt Book Courses Jane Sorry gt Book Timeslots gt Negotiate Timeslots gt My Schedule b View Class Timetables gt View Course Timetable b Log Out You need to Ist book the time slots inorder to negotiate the timeslots Back to Previous Page If the user books the slots and later requests to negotiate the time slots permission is granted to them Below is an illustration of this process Jane J W100 Negotiate Time Slots List of Slots you d like to negotiate gt My Schedule Slot ID Course ID Year View Class Timetables gt View Course Timetable Log Out Yearl Year 2 Mon hu ENN 5 We me RB sew so 3 45pm Slot M3 Slot Slot TH3 Slot mme sermon C aiaa Non negotiable IN eogotiable 52 5 6 3 System Testing The overall function of the system was examined and tested The results were examined to determine if they were correct The correctness of the system was tested by comparing the end result timetables with previously manually generated Based on this comparison it was concluded that the automated system saves a lot of time and energy spent on the generation of timetables manually The sy
30. ethod o aSkill stereotyped method can only be called during the decision phase of an agent Aptitude Aptitudes show the ability of an agent to reason both about knowledge and beliefs it owns An aptitude stereotyped attribute represents a functioning datum or parameter of a reasoning made by an aptitude stereotyped method For example an integer value represents the exploration depth of a planning tree A method stereotyped with aptitude expresses a reasoning which an agent Is able to do using its perceptions and its world representations This method uses one or several parameters that are attributes also stereotyped with aptitude For example a method allowing actions planning decision mechanism A method or an attribute which is stereotyped with aptitude can only be accessed affected or called by the agent itself to express its autonomy of decision Rules o an attribute or a method that is stereotyped with aptitude is necessarily private o aptitude attribute can only be used by a method that is also stereotyped with aptitude o amethod which is stereotyped with aptitude can only be called during the decision phase of the agent 21 anaptitude stereotyped method only call methods attributes that are stereotyped with perception representation interaction for instance for messages Representation World representations of an agent or beliefs concern other agents th
31. f the system where the various models of the system are developed Finally the system is tested in various capacities for any errors that could have been made during the implementation stage Unit testing will be done whereby each component of the system is tested separately then slowly integrated to see whether the system is able to operate as a single unit 26 Reasons for choosing this methodology v This methodology was specifically designed foradaptive multi agent systems based on AMAS theory ADELFE provides a new methodology to design a society of agents exhibiting a coherent activity v ADELFE has been used or is used in several case studies an intranet system design timetabling problem a flood forecast system in progress a mechanical design system in progress and a bioinformatics system in progress hence it passes of as a best practice v ADELFE is aimed to be a development toolkit of software with emerging functionalities and not only a mere methodology Therefore ADELFE will be able to provide some tools and libraries to ease the design and development of AM AS systems v ADELFE uses UML and AUM L notations which makes it easier to understand the system and one is able to quickly identify all application deliverables therefore the system developers can significantly improve project plans and track real progress 27 CHAPTER 4 ANALYSIS AND DESIGN 4 1 PRELIMINARY ANALYSIS The means to which prel
32. g this method to generate timetables How long does it take to come up with a timetable manually What are the things you put into consideration when coming up with a time table What would be the hard constraints when coming up with timetables How exactly do you come up with a timetable The procedures and steps you follow so as to come up with a timetable What are the problems that you encounter when coming up with a timetables What recommendations would you have for developing such a system 57 APPENDIX PROJECT PLANNING AND RESOURCES PROJECT PLAN 1 Preliminary Requirements Requirements ommo mom ME Nov 2011 Dec 2011 Jan 2012 Feb 2012 Mar 2012 Apr 2012 Start Finish Duration 23 10 30 10 6 11 13 111 20 11 27 11 4 12 11 12 18 12 25 12 15 1 29 1 29 4 5 Implementation Coding 3 2 2012 4 12 2012 20 em a PROJECT RESULTS AND PRODUCTS An online system that will be able to generate feasible timetables with respect to various constraints A multi agent system for scheduling sessions A user friendly interface A user guide on how to operate the website RESOURCES REQUIRED Expendable resources Printing paper Hardware Internet access Software for web hosting such as apache web server and dream weaver for developing the website Java Script and PHP to manage the constraints Informants for the university timetabling process Previously published research
33. h problems are not always 100 compatible for the university According to Abdennadher and Mohamed 2007 the number of courses academic resources room resources and university specific requirements are all factors which make the problem variant and hard Hence the attempt made by universities to generate timetables that best suits their needs is becomes difficult and complex as the number of constrains increase It goes without say that the generation of functional timetables requires a lot of time and effort In addition manually generated schedules usually fail to satisfy all essential constraints which are critical for the operation of the university Thus automating the generation of course timetables is still a challenging task according to McCollum 2006 2 2 NP COMPLETE PROBLEMS According to Agerbeck and Hansen 2008 Problems that can be solved by algorithms in polynomial time are considered to be easy problems For a problem of size n the time needed to find a solution is a polynomial function of n Harder problems require an exponential function of n which of course means that the execution time grows much faster than for an easy problem when the size of the problem increases NP complete problems are hard problems to solve They belongto a class of computational problems for which no deterministic polynomial algorithm has been found NP complete problems are a subset of the class NP Non Deterministic Polynomial A Non determi
34. hese tasks using the side bar menu The system also provides users with an interface that allows users to enter data from remote areas There are various forms to be filled in this system these forms are processed and validated and processed on the client side and then this information Is sent to the server Examples are shown below Lecturer registration form 71 Student Registration Form Lecturer Registration _ PT Bmhdeinmeliglstt 72 5 1 Slim Abdennadher Mohamed Aly 2007 Constraint Based University Timetabling for the German University in Cairo http met guc edu eg Repository Faculty Publications 67 Paper pdf accessed on 8 September 2011 2 Yan Yang Raman Paranjape LuigiBenedicenti 2006 An Agent Based General Solution M odel For the Course Timetabling Problem http jmvidal cse sc edu library AAM AS 06 docs P3fs405 pdf accessed 19 October 2011 3 Bernon Gleizes S Peyruqueou Picard 2007 ADELFE a M ethodology for Adaptive M ulti Agent Systems Engineering http www agent ai doc upload 200307 bern02_1 pdf accessed on 25 July 2011 4 M cCollum B Ireland N 2006 University timetabling Bridging the gap between re search and practice In Proceedings of the 6th International Conference on the Practice and Theory of Automated Timetabling 5 C Agerbeck Hansen 2008 A Multi Agent Approach to Solving NP Complete Problems http ci
35. icated timetable in maximum 5 20 minutes For simpler timetables it may take a shorter time under 5 minutes in some cases a matter of seconds For extremely difficult timetables it may take a longer time a matter of hours Mimosa Scheduling Software Mimosa is a Finnish scheduling and course planning software for use in any kind of school and university of varying type and size Besides academic timetabling M imosa is used to schedule conferences and in business and organisations M imosa has a very rich set of efficient optimisation tools and interactive timetabling selections even for the most demanding scheduling challenges M imosa also uses M ulti sensory approach using the Microsoft Agent character Peedy 2 9 CHAPTER SUMMARY In conclusion the generation of a feasible timetable in any educational institutions is critical for the proper management and utilization of school resources Therefore the automation of the timetabling process is necessary so as to ensure the generation of feasible timetables and one way of doing this is by using the agent based approach whereby a multi agent system M AS is used The use of MAS would be ideal due to the fact that the course timetabling problem is considered to be NP complete problem and the idea of using a M AS can be used as intuitive approach The MAS uses negotiating agents also known as multi agents to reach compromises between conflicting interests by using the active mode principles
36. ilities and human resources 30 2 2 SKSTEMTREQUIBEM EN TS udin 31 4 2 1 Characterization or bd ditte 31 2 29 5tem ENU E 31 4 2 e a A 31 12 4 Determination Ofthe 36 4 2 5 Identification of co operation failures mms 37 2 5 DESIGN An durch Um sip aae 38 45221 elass identi 38 Tie IDterclass TelatloriSIDSsus 39 4 3 3 The preliminary class pl b D Padre t nad br cl n 40 4 3 4 Verification of the global level AM AS adequacy 40 4 8 5 Identification Of aae li S asser PED ______ _ _ 40 4 3 6 Agent Interaction with other nenne 4 4 3 7 Verification of the local level AM AS 4 4 3 7 Detailed Architecture and Agent Model 42 Agent Al tide Ma cd eR 43 4 3 9 Complete design Cla Gla 45
37. iminary requirements were gathered was through interviews The interviews were carried out with the people charge of creating timetables the school Through these interviews the key requirements of the system and also techniques that would deem key to the development of this system were attained The interviews were scheduled based on the interviewees availability A set of questions was used to guide the interview and are available in APPENDIX I 4 1 1 Results of the interview The following was noted after carrying out the interview The current system The university over the years has been generating timetables manually This is done in meetings whereby lecturers within a school sit down and distribute various courses to different lecturers Once that is done each lecturer is given a time slot in which they are comfortable teaching at that time In case of conflicts there is need for negotiation between lecturers Conflicts take a substantial amount of time to resolve depending on how flexible the lecturers schedule is The key stakeholders for the course timetabling problem are i The lecturers Lecturers have some constraints about their availability whereby each of them have different time preferences in which the can teach a particular lecture Another constraint would be the lecturers capability to teach a certain course whereby a lecturer is assigned a course in which he she is comfortable and passionate about teaching
38. ing System One needs to have installed Xampp server on their computer To launch the web based application first start the Xampp server You may find the start up icon on your desktop or you can go to Programs and start it up from there Open the browser of your choice and specify the local host address which directs you to where your application is stored A user has to present his her identification details together with a password in order to be authorized to access and utilize the timetabling system There are various links on the home page that allow a user to perform various tasks Such tasks and access to information depends on the administration rights a user has Information is limited to ordinary system users while administrative user has rights to access all the information in the system There are three different user levels The student lecturer and administrator user The student user login leadsto a group of pages that allow the user to register and view the timetables Registration of courses ABOUT US ARTICLES CONTACT US SITE MAP d 4 gt 3 S gt 14 s HOME A gt Y gt View View Class Timetables Welcome Stephanie L STUDENTS Register for courses Registration Number Degree Name School Of Mathematics ear 1 Semester 1st Semester Register This is a password protected area only accessible to members of this site Back to Previou
39. ises among agents with conflicting interests Such a system will significantly reduce the amount of time and energy spent ingenerating timetables and at the same time optimize on all the available resources as a result of using agents Consequently the utilization of agents makes the system flexible whereby it is able to adapt to different scenarios and behave in an appropriate manner and also the fact that it s web based makes it easily accessible to its users Table of Contents ECE RATIO 1 ia qe AN Cs s new 2 0 3 LST OF ABBREVIA HON c 7 CHAPTER T MIN ROT IN nin E E 8 LE BGR RUIN eera E E 8 EET I Es NIME TRE OSTSEE 8 1 1 2 Timetabling in the nnns 9 1 2 PROB ETC 10 L3 OBE CM VES OFTHE STUDY verb cesi ame usan 11 LX SESTEM ome iab PUO 11 12 CHAPTER 2 LITERATURE REVIEW urne eru ceat NAM 13 13 13 2 3 M ULTI AGENTS SYSTEMS 14 The scheme of the
40. ith minimal message passing The scheme of the scheduling MAS conflict group Z2 84 t 4 CENTRAL AGENT Gonrliot group 1 1 conflict group 3 14 2 4 AN AGENT SOLUTION MODEL According to Yan Raman and Luigi 2006 all course timetabling problems have certain common hard constraints I e 1 No student has more than one course at the same time 2 Only one course is scheduled in a classroom at any one time 3 The classroom assigned to any course is large enough to hold all enrolled students By identifying the common hard constraints provides the basis to establish a general solution model for the course timetabling problem using agent technology course is an event involving course name class time classroom and people i e students lecturers and teaching assistants There are a set of constraints which govern the relationships between these entities In the proposed agent solution model each agent represents a constraint in the course timetabling problem The model includes the following agents v The CHC Agent which represents the Common Hard Constraints identified above v The HC Agents which represent other additional Hard Constraints such as no class scheduled during the lunch break time v The SC Agents which represent the Soft Constraints such as no student assigned more than two dasses
41. le Lecturer ID 1 Course ID getCourse ID Course Manager Course allocation ID Constraint Constraint ID Constraint Type Body of the constraint getLecturer that can teach the course get Lecturer ID USES Lecturer Time slot ID Room allocatioin ID Course allocation ID Timeslot preference 1 Timeslot preference 2 Timeslot preference 3 Student ID Degree Academic Year getacademic year USES Room ID Nature LI Course Nature Degree assosiacted with course Academic Year in which couse is done Lecturer that can teach the course Number of students doing the course USES Room Allocation ID 1 Room ID Course ID getNumber of Students undertaking the course getCourseNature getRoomcapacity getRoomnature 39 4 3 3 The preliminary class diagram Attribube courses 1 Define constraints Attribube courses Occupies 4 3 4 Verification of the global level AMAS adequacy At the global level we need to answer the question is a system implementation using AM AS needed For the case study the decision tool clearly suggests to use the AMAS to design the global level Moreover the tool indicates that some entities could be decomposed as AS So once the agents are Identified there is need to reuse the method on them 4 3 5 Identifica
42. m the form Prevents SQL injection function clean str 4 trim str if get magic quotes gpc 1 stripslashes str j returnmysql real escape string str Sanitize the POST values mode POST ogin username POST username password clean POST password Input Validations if mode errmsg arr Please choose who your logging in as true if username lt lt errmsg_arr Login ID missing true 62 if password 4 errmsg arr Password missing errflag true f there are input validations redirect back to the login form if Serrflag SESSION ERRM SG errmsg_arr session write close header location login php exit j 11 Check whether the query was successful or not if mode student Create query qry SELECT FROM studusers WHERE username username AND password md5 POST password result2mysql query qry if result if mysql_ num_rows result 1 1 Login Successful session regenerate 10 member mysql_fetch_assoc result SESSION SESS M EM 104 member studmemID SESSION SESS FIRST 2 member fn 63 SESSION SESS LAST member In SESSION studid member studid session write close header location member index1 php exit lelse Login failed hea
43. mportant or critical Such constraints are known as soft constraints that should be satisfied but maybe violated Lecturers can be assigned more than one courses and furthermore some constraints can be taken into consideration to improve the quality of education e g No lectures should be scheduled in the last slot of any day or a certain day should have an activity slot for students and a slot where all university academics can meet For those slots no sessions should be scheduled In summary the timetabling problem is too complex since a lot of constraints should be taken into account when scheduling a session Due to the complexity of the problem a good model is needed to solve the problem and this can be done through the help of a multi agents that are able to compromise conflicting interests to come up with a feasible timetable This would in turn save weeks of manual work 1 2 PROBLEM DEFINITION In the old days schools taught a limited number of compulsory subjects The number of students as well as teachers was small and the facilities within a school environment were limited In contrast to old days the modern world has advanced at a great pace in terms of population facilities and technology With the advancement of the mentioned areas the number of students and teachers has increased schools have expanded more subjects are introduced and more facilities are required to accommodate the modern education Therefore there is aneed t
44. multaneously in the best possible manner The timing of events must be such that nobody has more than one event at the same time With respect to the above considerations the process of coming up with a timetable for any learning institution is considered as a time consuming and interactive process that involves a lot of compromising on both the human and technical resources offered by the university The technical resources would be the class rooms and the human resources would be the lecturers and the students Hence university course timetabling problems are combinational problems which consist of scheduling a set of courses within a given number of class rooms and time periods The diagram below illustrates the relation between concepts of timetabling A Lesson In order to come up with a feasible timetable it is important to compromise between all lecturers time preferences all also keeping in mind availability of classrooms and students number of students and curricula Subject 1 12 Timetabling in University The university comprises of a number of schools all located in different campuses Each and every school has its own timetable and over the years the university has been generating these timetables manually This process is time consuming and the end result i e the timetables generated may or may not be satisfactory With respect to the project scope it would be ideal to zero in
45. ng actions perceive decide act To ensure that this method does exist an agent inherits from a super class called Cooperative Agent Rule o Anagent stereotyped class inherits from the Cooperative Agent class 19 Characteristic A characteristic is an intrinsic or physical property of an agent An attribute which is stereotyped with characteristic expresses the value of such a property For example the size of an agent the number of legs of an ant A characteristic stereotyped method is a means to modify or update an agent property For example a method that enables to modify the number of legs of an ant A characteristic can be called during any phase of the lifecycle of the agent A characteristic can be accessed or not by other agents Perception A perception is a means to receive information from the physical or social other agents environment An attribute perception represents a datum coming from the environment and is necessarily private For instance the number of agents seen by an agent at a given time a messages list A method stereotyped with perception is a means to modify or update attributes that are stereotyped perception This kind of method is not necessarily private because other agents can call it to send a message for example For example a method that enables to put a message in an agent mailbox Rule o An attribute stereotyped with perception is necessarily private
46. nistic algorithm is able to find a correct solution but it is not always guaranteed The solution is found by making a series of guesses and the algorithm will only arrive at a correct solution if the right guesses are made along the way A problem is called NP if its solution can be found and verified by a non deterministic algorithm in polynomial time In this case the NP complete problem is scheduling the timetable and the idea of using a multi agent system M AS can be used as an intuitive approach The system will simulate the way humans think and interact to construct algorithms and heuristics that can solve the NP complete problem i e timetable scheduling 13 2 3 MULTI AGENTS SYSTEMS MAS According to Opreal 2006 Multi agent systems are concerned with coordinating behavior among a collection of autonomous intelligent agents that work in an environment Sometimes the intelligent agents are designed to reconcile their own interests with the constraints implied by other agents or they can opt to become selfish and reconcile their own interest with no regards to the other agents in the environment The complexity of multi agent systems is generally higher than that corresponding to conventional software systems and their success rely on properly designed and well tested subsystems Also in the particular case of timetable scheduling the M AS could find optimal or a sub optimal solution using mainly inter agent communication w
47. o come up with a method of using every teacher class and facility effectively while maintaining the educational standards and this can only be done using timetables Currently the university generates its timetables manually and this process takes a period of 3 4 weeks This process has proven to be difficult over the years since it s generated by a group of lecturers who need to go back and forth so as to come up with a feasible timetable Some of the issues that arise from generating the timetable manually is that a lot of effort is used in creating new instances of a timetable whenever a conflict arises and this in turn wastes a lot of time and energy Another issue that arises is coming up with a feasible timetable that will result in utilizing the highest productivity of teachers optimum use of facilities and classrooms while maintaining the required number of timeslots for the respectful subjects Often the timetable generated prove not to be satisfactory hence not feasible One of the objectives is to take advantage of the latest technological advancements to come up with a timetabling system that is able to tackle the above Issues by using multi agents that are able to compromise conflicting interests to come up with a feasible timetable in the shortest time possible 10 1 3 OBJECTIVES OF THE STUDY The main goal of the system is to e Provide the university with a web based time tabling assistant that is able to generate a timeta
48. reotyped with skill or interaction 22 Co operation The social attitude of agents is implemented using rules allowing Non Cooperative Situations NCS solving An agent must have a set of rules predicates that allow it to detect NCS It must also have a method to enable it to solve NCS this method associates actions with situations in order to process them A method that is stereotyped with cooperation is always private and can be of two kinds o amethod that returns a boolean result to detect a situation its parameters are perceptions representations or skills o asolving method a priori one per agent that allows the association of one several solving actions with each situation These methods are called only during the decision phase of an agent Rule o Acooperation stereotyped method is private 2 7 STATISTICS The multi agent approach to the university course timetabling problem has proven successful in offline experiments We believe that its major advantages in comparison to the other approaches of timetabling are v It is aimed directly towards the resolution of the conflict spots in the timetable assuring a relatively good efficiency Itallows different strategies and preferences to be employed within a well defined negotiation framework tisanatural computational model for some types of scheduling problems that can be projected onto a multi agent system in a straightforward manner 2 8RELATE
49. rtners to follow or to give each course These goals raise the problem of ubiquity Agents cannot be at different places at different moments Therefore we propose to create one agent per course for each teacher or student group Two agent levels are distinguished RepresentativeAgent at the highest level it represents a teacher or a student group within the system BookingAgent BA at the lowest level it is responsible for finding partners to negotiate with in terms of time slots and booking rooms for a RA There are as many BA as the number of courses a teacher has to give or a student group has to follow 4 4 3 7 Detailed Architecture and Agent Model This phrase of the ADELFE involves identification of software components and their descriptions in terms of blocks classes agents and interactions thus making the architecture of the system The agent model is also included in this architecture In this case the design pattern would be named co operative agent architecture and it has the following four packages Agent package to manage BAs and RAs Grid package to manage the different dimensions of a grid and its cells Constraint package which has to be accessible to rooms and agents Interface package to enable a user to interact with the system 4 3 7 1 Package diagram Constraint Package Interface Package Package Agent Package 42 4 3 7 2 Agent Model Diagram Database
50. s You ve Successfully Booked Courses Negotiate Timeslots My Schedule View Class Timetables View Course Timetable Log Out You ve successfully booked your courses You need to book time slots Please book timeslots for the courses that you ve choosen 50 The following message 15 displayed when a lecturer has successfully booked timeslots for their respective courses ABOUT US ARTICLES CONTACTUS SITE MAP Book Courses gt Book Timeslots You ve Successfully Booked Time Slots gt Negotiate Timeslots gt My Schedule View Class Timetables View Course Timetable 22 BN Ya Log Out You ve booked slots for your courses You can now negotiate time slots if interested The following message is displayed when a system user has successfully logged out of the system P HOME _ ABOUT US ARTICLES CONTACTUS SITE MAP You have been logged out Click here to Login 51 5 6 2 Integration T esting Results of functions were passed to each other where applicable For example in order for you to negotiate for a time slot you first need to book a timeslot This is because the lecturer user needs to trade in slots so as to participate in the negotiation process If a user tries to negotiate for timeslots before booking the time slots an error message 15 displayed instructing the us
51. s Page 67 View timetables gt View View Class Timetables gt Log Out Welcome Stephanie STUDENTS SCI Timetables Year Wed Fri Year 2 Wed Parcas 52212 ne 2 00pm rcp Cum scia _ The lecturer login leads a group of pages that allow the lecturer to book courses time slots negotiate timeslots view their schedule and other timetables as well Booking Courses 4 SER jd W Ww HOME ABOUT US ARTICLES CONTACTUS SITE MAP gt Book Courses Stephanie Book Timeslots SCHOOL OF COMPUTING AND INFORMATICS b Negotiate Timeslots b My Schedule gt View Class Timetables gt View Course Timetable gt Log Out Booking Courses Lecturer Ib bourse 1 to Programming Y ourse 2 se Introduction to Programming Y BOOK COURSES 68 Booking Timeslots gt Book Timeslots gt Negotiate Timeslots b My Schedule gt View Class Timetables gt View Course Timetable Log Out Negotiating Timeslots gt Book Courses gt Book Timeslots gt Negotiate Timeslots gt My Schedule Book Time Slots For each course book two slots Course Course Name Mear CSCAM Computer Network Security 4 5 4141 ICTs and Socie 4 1 Slot Availability
52. s not Satisfactory for them Another way of improving the system is to incorporate a prioritization mechanism whereby high level users are given first priority to access and use the system 6 5 Conclusion Using the web application developed the school is able to generate timetables with the greatest of ease The automation of this system has greatly reduced cases of conflict and 15 able to create timetables from scratch according to user preference and provides a negotiating platform in case a user is not satisfied with the time slot presented to them For the potential users of the system providing access to such a facility saves a lot of time spent in the generation of timetables Also the system is able to come up with feasible timetables on time using the least amount of energy whereas significantly reducing the number of conflicts Implementing such a facility in the university will prove to be an added advantage for it will save up a lot of time and significantly cut down on any expenses used during the generation of timetables Since the system 15 web based it will be accessible to everyone and also the expenses on web hosting are significantly low hence more reason to implement the system 56 APPENDIX Automatic Timetabling System Interview Questions p4 How does the university currently handle the generation of timetables M anual Semi automatic automatic How long has the university been usin
53. ss majority of the users needs and generate a feasible timetables 25 3 3 METHODOLOGY The methodology that was incorporated in the system development wasthe ADELFE methodology ADELFE is the French acronym for toolkit to develop software with emergent functionality Atelier pour le DEveloppement de Logiciels FonctionnaliteEmergente 3 3 1ADELFE Overview The ADELFE process consists of six work definitions Preliminary Requirements The preliminary requirements work definition mainly deals with the description of the system and the environment in which the system will be deployed It consists of defining the most appropriate system for end users Final Requirements The aim of the final requirements is to transform this view in a use case model and to organize and to manage the requirements functional or not and their priorities Analysis this involves the identification of the agents The analysis work definition has to develop an understanding of the system its structure in terms of components and to know whether or not the AM AS theory is required Design The design work definition aims to formulate models that focus on non functional requirements and the solution domain and that prepare for the implementation and test of the system In ADELFE agents being identified and their relationships being studied designers have now to study the way in which the agents are going to interact After design follows the implementation o
54. stem was also found to be flexible for it has a negotiation platform Below are some examples of the automatically generated timetables based on the environment constraints Year 4 har Lecturer Lecturer ID Course Name ID First name Last name 5 411 Computer Network Securi SWIOO Stephanie rinciples of Business 5 413 and 100 Entrepreneurshi 50414 ICTs and Socie SWIOO Stephanie 5 416 Computer Systems Project scar Eniroductiontelanwege LII F Fri fam mw Eee Back to Previous Page 53 5 7 Security All users who access this system must be authenticated first The authentication method used this system 15 the use of a username and a password A log in page was created for registered users i e lecturers students and administrators to access the system A sample of this is shown E X 2 gt Chap Chap TIMETABLING i de SA ABOUT US ARTICLES 5 SITE MAP gt Student Log As 7 Lecturer 7 Administrator Username a Password Reset Log In I m not a member but i d like to Register 54 CHAPTER 6 DISCUSSION 6 1 Achievements Several achievements were made as a result of undertaking this project These include Development of a system that will help in the automatic generation of feasible
55. straint agents passively receive and evaluate proposals For this reason we call this the passive mode 2 4 2 Active Mode In the active each constraint agent has connections to all of others In addition there is a token which is passed around between the constraint agents Only the agent owning the token has the right to make proposalsto try to improve the course timetable and the proposals are guaranteed not to violate its own constraint If a proposal violates the common hard constraints it will be rejected because all timetables must always be basic feasible course timetable even if they do not reduce other constraint violations Otherwise the timetabling process is similar to that of passive mode In the active mode all the agents equally contribute to improve the course timetable based on their own constraints This allows the constraint violations to be reduced more effectively and efficiently That is also the reason it is called the active mode If a solution cannot be found after the token has been passed for a certain number of rounds among all the agents the timetabling process is ended hence can t guarantee an optimal solution satisfying all the constraints 17 HC Agent 2 4 3 Parallel active Mode The active mode can be further modified to make the token unnecessary allowing all the constraint agents to work in parallel In this situation the CHC Agent acts as a mediator to judge if the proposals are valid or not
56. taken if T assign course Return capability H f F not qualified to teach this course Course request declined not qualified 34 Room manager entity interaction with the system Package Room manager Request for room stating the course and time slot nature of cou rse theoretical practical if practical check lab availability in that time slot Result if F change time slot ii student capacity if no of students lt room capacity assign room Use diffrent time slot if T check number of students doing that Return student capacity Book room Change time slot Room request granted no of students gt room capacity split student group Room big enough consider splitting student group ifthoretical check lecture room availability in that slc if F change time slot Use diffrent time slot if T check number of students doing that course student capacity if no of students lt room capacity assign room Book room Room r
57. tem Database precondition Number of students taking the course should match the room capacity 31 Sequence diagram between system and active entities Student entity interaction with the system Package Student Enter username amp password Validate access T F Log in if F register lecturer Present registration form invalid username or password if T show timetable Show timetable Current timetable Lecturer entity interaction with the system Top Packaae Lecturer Enter username 4 password Validate access T F if F register lecturer Present registration form if T prompt request Time slot prompt Present time slot appllication form Time slot request Check availability of time slot if T allocate time slot Time slot request granted if N modify time slot request Change time slot request Time slot not available Course manager entity interaction with the system Top Package Course manager Request for course Check availabilty T F if F course already taken Change course Course request granted course if T check capability to teach course Course already
58. teseerx ist psu edu viewdoc download doi 10 1 1 93 4736 amp rep rep1 amp type odf accessed on 5 January 2011 6 M Oprea1 2006 M ulti Agent System for University Course Timetable Scheduling http fmi unibuc ro cniv 2006 disc icvl documente pdf tech 2 oprea pdf acce ssed on 27 September 2011 13
59. timetables for the university Appreciation of web based applications solutions A better understanding of software development techniques and design and a wider knowledge of computer technologies Development of a system that has a significant impact on real world application A basis for further software development work 6 2 Constraints The main challenges encountered were Limited Information about multi agent systems e Limited examples of multi agent systems e Program incompatibility with agent based programs in terms of web hosting 6 3 Recommendations The following are additions that can be made to this project e Development of the system as a mobile application e Incorporating the system into the university as a whole so as to incorporate all the schools within the university e Send alerts via sms to users when any changes are made on the timetable e Allow prioritization of users such that high level users are given first priority in booking courses and timeslots 55 6 4 Suggestions for Further Work One way of improving the web application is to incorporate an emailing trigger whereby the timetables could be emailed to the users with respect to their associatively This would ensure that every user gets a copy of the timetable Another way of improving the system is by creating another negotiating platform for courses This would allow lecturers to exchange courses if the choice that they got wa
60. tion of agents In this phrase of the ADELFE methodology we need to identify the co operative agents the system These agents are simply entities that ignore the global function of the system such that they pursue an individual objective and try to be permanently cooperative with other agents involved in the system At this stage we identify lecturers and students groups as being cooperative agents All other entities are considered as objects Teachers and students are autonomous have local views are plunged in the rooms and have to negotiate to find partners and to resolve resource problems 40 4 3 6 Agent Interaction with other entities Attribube courses 1 lt lt Co operative agent gt gt Lecturer Manages constraints Define constr Room Manager Occupies Attribube courses Co operative Agent Student Occupies Three classes of agents appear in the above diagram The StudentsGroup the Lecturer and the BookingAgent The two firsts are interface agents between the system and the users Booking Agents aim to reserve time slots Cell of a Grid and Rooms for Teachers or StudentsGroups in terms of their Constraints 4 3 7 Verification of the local level AMAS adequacy If the first step of adequacy to the AMAS theory indicates a possible decomposition each agent has to be analyzed as a system The goals of an agent Lecturer or Student are to find different places and pa
61. up other courses A student group not able to get a room within a given time slot A room is not big enough to accommodate a certain student group 37 4 3 DESIGN 4 3 1 Class identification The classes that are present within the system would be The lecturer class the class represents all the lecturers within the university and their time preferences in terms of their availability The student class the class represents all students groups within the university based on their major The room manager class the class represents the allocation of courses to various rooms based on the room capacity and the number of students undertaking that course The course manager class the class represents the assignment of course to lecturers based on their capability to teach The room class this class represents the all the rooms within the university and their corresponding capacities and sole purpose lab lecture room The database class this class represents all the information one would want to know regarding a certain course The constraint class this class represents the constraints that are declared by each entity The constraint manager class it represents the manner to which the constraints will be handled and executed The cell class this class represents the various intersections in terms of time slots The grid class this class is used to visualize the current timetable 38 4 3 2 The interclass relationships timetab
62. web tools used in this project 5 4 1 PHP PHP is officially known as PHP Hypertext Pre processor It is a server side scripting language that is often written in a HTML context PHP code that has embedded HTM L fragments in it is not sent to the client but is interpreted on the server side and the resultant HTM L is combined with initial HTML fragments and then sent to the client PHP was selected for the following reasons a Itisan excellent tool for server side scripting and allows for easier automation that was traditionally handled by Perl or Shell scripting b It addresses many of the web programming issues that programmers have faced while using languages such as C and Perl PHP has a rich library of functions to handle almost everything when developing for the web c PHP performs well since a programmer need not rely on other utility libraries that often slow things down d PHP also supports object oriented programming which makes the final code easier to debug allows code reuse during development and lends itself well to the MVC M odel View Controller architecture This shortens the development time considerably e PHP isalso very easy to configure on a web server and most all if not all web browsers support it PHP is well documented and user forums are widely available on the internet 4 f PHP does not put strain on servers It uses its own inbuilt memory space that decreases the workload from the servers and the
Download Pdf Manuals
Related Search