Sandbox‎ > ‎Archive‎ > ‎IPT 2008-09‎ > ‎Jason‎ > ‎

Jason's updates

10 September 2009 (Thursday)

posted Sep 9, 2009, 9:37 PM by Unknown user   [ updated Sep 9, 2009, 9:44 PM by Eddie Woo ]

08 September 2009 (Tuesday)

posted Sep 7, 2009, 3:26 PM by Unknown user   [ updated Sep 22, 2009, 5:36 AM by Eddie Woo ]

HSC Trial Exam Review

Final mark: 19 + 83 = 102/125

Section A: Multiple Choice
  • 1-5. Full marks.
  • 6. Query by Example (QBE) would have been a suitable search method. However, no-one in the class knew this (due to the content not being taught) so the question will apparently not be counted in the final mark.
  • 7-20. Full marks.
Section B: Extended Answer
Question 21:
  • a) Full marks – although one sentence I added on at the end should have been considered as part of the first issue. Also, important to remember that cost is a social issue that works for just about any situation (just as privacy is a recurring ethical issue).
  • b) Full marks.
  • c) Full marks.
  • d) Full marks. 
Question 22:
  • a) i) Lost 2 marks – each 24-bit colour pixel only requires 24 bits of data to be described, not 2^24 bits (24-bit colour can display 2^24 different colours). I remembered to divide by 8 to convert bits into bytes, however I then wrongly divided by 1000 (instead of 1024) to convert bytes into kilobytes.
  • a) ii) Lost 1 mark – did not relate my answer to the particular image. Need to remember that the GIF format can only handle a maximum of 8-bit colour, and that it is no less suited to high resolution images than lower resolution images.
  • b) Full marks.
  • c) Full marks, however need to remember to link it back to the given scenario/situation.
  • d) Lost 2 marks – I identified many appropriate examples of developments in technology, but needed to further discuss the issues and implications associated with them.
Question 23:
  • a) Lost 1 mark – did not adequately relate one of my identified issues to the iPhone.
  • b) i) Lost 2 marks – need to identify more entities and processes.
  • b) ii) Lost 2 marks – the HSC keyword “critically analyse” requires relationships to be drawn between the issues and elaborated in a way that illustrates a depth of understanding. I identified issues in each of the key named areas, but did not elaborate on the relationships between them.
  • c) Full marks. 
Question 24:
  • a) i) Lost 1 mark – the conditions in my decision tree were not sufficiently distinct and needed to be more related to information provided in the question, however the overall layout of the decision tree was correct.
  • a) ii) Full marks – identified the correct situation and justified it adequately.
  • a) iii) Lost 1 mark – need to include further methods of analysis used by the system.
  • b) Lost 3 marks – the HSC keyword “discuss” requires that both positive and negative sides be considered; my response was very one-sided. 
Question 25:
  • a) i) Lost 2 marks – needed to relate my response more directly to the given scenario.
  • a) ii) Full marks.
  • b) Full marks.
Question 26:
  • a) Lost 1 mark – did not read the article thoroughly enough, forgot to include international phone calls as a type of technology being restricted.
  • b) Lost 1 mark – part of my response did not answer the question. Also need to distinguish between events in the article that were alleged and officially confirmed.
  • c) Full marks, but need to avoid using colloquial language in an exam situation.
  • d) Lost 1 mark – needed to write a bit more (considering it was a 5 mark question) and relate my argument more directly back to the given scenario.
  • e) Lost 2 marks – the HSC keywords “compare and contrast” requires discussion of both similarities and differences, my response only covered differences between computer based and non-computer-based communications systems.

01 September 2009 (Tuesday)

posted Aug 31, 2009, 8:56 PM by Unknown user   [ updated Sep 9, 2009, 9:39 PM by Eddie Woo ]

From Jason and Kevin:
 
2006 HSC paper q21:
 
a) SELECT CountryName
FROM Countries
WHERE Continent="South America" AND Language="Spanish"
 
b) (200 x (25 + 15 + 6 + 6 + 6 + 10)) / 1024

20 August 2009 (Thursday)

posted Aug 19, 2009, 4:53 PM by Unknown user   [ updated Aug 19, 2009, 6:07 PM by Eddie Woo ]

2008 HSC Paper
 
Section I: Multiple Choice
  1. A
  2. D
  3. A
  4. D
  5. A
  6. B
  7. B
  8. C
  9. D
  10. B
  11. C
  12. D
  13. D
  14. C
  15. C
  16. A
  17. C
  18. C
  19. B
  20. B

20/20

Section II
21.
a) see context diagram in book 3/3
b) "Owner_Telephone_Number" should be text because it may contain non-numerical data (e.g. dashes, brackets). "Paid" should be a boolean because its value will always be either yes or no. 3/3
c) May require a new table in the database to be creatd for vet details. Entire PETS system can be converted into a relational database; each pet record can have several relationship fields linking to the appropriate vet's details.3/4
 
22.
a) Staff will need to know how to: scan student smart cards, generate late slips, request a SMS/email to be sent, log in to and use the web portal to generate attendance reports. Staff will be trained through documentation supplied by the system designers as well as using the actual system with test data. Training will need to take place through most of system development and after the system has been implemented; i.e. from making decisions and designing solutions (consult with users) right through to implementation and testing.
b) After the initial infrastructure for the system has been set up, network administrator is responsible for its ongoing maintenance. Distributing smart cards to individual students, replacing those that have been lost. Managing accounts (levels of access), backup and security of stored data, testing and evaluation.
c) Data integrity means ensuring the stored data is correct and has not been corrupted or tampered with; user accounts and levels of acccess to prevent unauthorised changes; incorrect data = system cannot function effectively. Must prevent unauthorised use of messaging system, ensure that the system only sends messages when necessary and to the right people.
 
23.
a) Questionnaire/survey - allows people to anonymously suggest features that they want and suggest ideas for site design. Interviews - can obtain more in-depth information from potential users; however may be intimidating.
b) Facilities for sharing photos/videos are clear; however unsure how text-based messages are to be exchanged. Maybe add a new tab up the top specifically for posting messages, allow users to "communicate with ease".
c) Access - ensure that only authorised users can access their own profile (use of user accounts, levels of access for users and administrators), also have to determine who is allowed to access/modify other people's data. Ownership and control of data - who owns the data supplied by users, system administrators have responsibility to ensure inappropriate/illegal/unethical data is not exchanged using their site. Privacy principles - users must be able to control who can see their data, have a right to keep some personal details private.
 
24.
a) Satellites and other forms of wireless communication usually utilise radio/microwaves transmitted through the air, can cover a large distance. Cables are expensive to cover large distances but generally allow faster and more reliable transmission of data. For this system wireles would be more suitable - it is in a remote region, laying cables through trees and natural area would be impractical/expensive, satellite allows for transmission of data from base station to computers elsewhere (internationally).
 
Section III
 

11 August 2009 (Tuesday)

posted Aug 10, 2009, 3:38 PM by Unknown user   [ updated Aug 10, 2009, 3:46 PM by Eddie Woo ]

Data mining (pg. 277, Jacaranda)
 
23. Who is responsible for decisions made using DSSs? Explain your answer.
aaa
 
24. Summarise the process of building a DSS that involves an expert system. List the people involved and describe their roles.
aaa
 
27. Explain the role of input weights or "fuzzy logic" in artificial neural networks. What advantages does this give them over conventional systems?
aaa

04 August 2009 (Tuesday)

posted Aug 3, 2009, 3:47 PM by Unknown user   [ updated Aug 4, 2009, 8:03 AM by Eddie Woo ]

Knowledge Engineering

1. Identify some of the factors that make knowledge acquisition complicated.
Expert systems require the acquisition of knowledge possessed by human experts, which poses difficulties in that is often poorly organised, not explicitly expressed, or unstructured. This raw knowledge can also come from other locations such as paper or electronic documents, adding further complications - collected data comes in many kinds, many forms, and from many sources. Transferring knowledge from one person to another is often a difficult task - how does the human expert express all their knowledge and experience in a way that is understandable by the knowledge engineer? Even once the engineer fully understands the information involved (and has essentially become almost an expert themselves), they must find a suitable way to structure and accurately represent this within the knowledge base. A large number of participants can lead to further problems, especially where there are conflicting viewpoints or when irrelevant data is produced.

2. Evaluate the importance of knowledge validation in a decision support system's knowledge base.
As with any computer-based information system, validation and testing of data is vital to eliminate errors. Specifically designed test cases must be used to test data and that the system is functioning correctly. Fundamental errors and flaws in the knowledge base would result in an ultimately faulty expert system that is unable to make reliable decisions and find the optimum solution from a given problem or situation.

3. Describe a situation in which a decision support system a) requires and b) does not require the ability to provide sophisticated explanation and justification for its advice.
[do this later]

4. Using the Web as a knowledge collection tool overcomes some of the problems of data acquisition but creates others. What problems does it overcome and create? (Infobox, pg. 217)
[do this later]


Search Algorithms
 
Google
Ask
AltaVista
Yahoo
Bing
[do this later, internet keeps dropping out]


5.6: Other information processes and issues related to decision support - Activities (pg. 220, Jacaranda)
 
Remember
2. Explain the difference between a depth-first search and a breadth-first search.
A depth-first search follows each branch of a knowledge base down its entire length to the deepest node (problem category) before backtracking up the hierarchy and progressing to the next downwards branch. This process continues until a matching solution is found or until all branches and nodes have been searched. Depth-first searches are particularly efficient for a tree that has a large number of levels, or where the solution is very specific and buried quite deep in the hierarchy. A breadth-first search searches horizontally across each level of the knowledge base before descending downwards a step and repeating this search pattern. This method is especially efficient if the solution isat a relatively high level in the hierarchy, i.e. if it is quite a general and broad solution.
 
3. What are the two main ways of finding information on the Web (according to the textbook)?
One method of finding information on the Web is to use web directories (such as Yahoo!). These store web addresses in the form of a hierarchical knowledge base, with the user allowed to select a category and delve down into it until they reach a website of relevance. Users can also help improve and expand the database by adding relevant links to the appropriate section within the tree. The other main method of finding information online is to use a search engine. These maintain indexed databases of billions of separate websites, and search through them based on keywords entered by the user. Different search engines work in different ways, although most generally attempt to rank the results in order by factoring in overall popularity and relevance to the user.
 
Think
4. Explain what a false summit is in a hill-climbing algorithm. What would be the consequences of this false summit?
A hill-climbing algorithm works by actively deciding in which direction the search should be proceeding. From a given point, the algorithm attempts to reach the "summit" of the (metaphorical) mountain of data, where higher levels indicate increased relevance to the user. However, as this method only takes into account the immediately surrounding data, it is possible for the algorithm to choose to travel in one direction and eventually find a point where it is not possible to get any higher (i.e. closer to the goal). Yet this could possibly be a false summit - if the algorithm had by chance chosen to take a different direction earlier, it could have found a different local summit (the highest point in the immediate surrounding area) and assumed that was the optimum solution. The landscape being explored by the algorithm is similar to a mountain range, in that the peak of each mountain is only accessible from the base of that mountain - once the algorithm starts going up, it is only going to reach the peak of the mountain it is currently on. There is no way for the hill-climbing algorithm to know whether it has found the real summit, thus it may eventually return an inferior solution that is not the best available to meet the user's needs.

31 July 2009 (Friday)

posted Jul 30, 2009, 5:45 PM by Unknown user   [ updated Aug 3, 2009, 3:41 PM by Eddie Woo ]

Processing and Decision Support

1. Explain why processing is the central process of decision support.
Processing, the creation of new data from existing inputs, is a crucial part of decision support systems dealing with both structured and unstructured problems. The other information processes generally only serve a few key functions - rearranging the data so that it is more suitable for use by different systems, transferring data between these systems and displaying it on each, or adding additional describing information to the data (metadata; data about data). Only processing actually creates new information, which in the case of decision support systems is generally the optimum solution to a problem given a set of defined inputs and constraints. Thus, processing is the central process of decision support - without it, a DSS would not be able to come to an effective conclusion.

2. Describe a shell and outline the reasons why programmers might include one in their decision support system.
A shell is essentially used to encapsulate a DSS's technical inner workings, ultimately aimed at making it more user-friendly and understandable for the end user. Shells are also used to provide an interface for the user to interact with - they accept commands, process them, and pass them on to another program or system. For programmers, a decision support system shell makes it easier for them to create the underlying expert system and populate it with knowledge relevant to the situation. Rules and existing knowledge can be entered without explicit use of a specific programming language.

3. Contrast the decision support that is provided by (a) a spreadsheet with (b) an inference engine.
Spreadsheets are largely user-driven, and rely on rules and calculations that have been defined and entered by the end user. Thus, they mainly provide decision support through forward chaining - starting with the known facts, the spreadsheet works through inputted rules to arrive at an appropriate conclusion. In this way, users cannot specify the eventual outcomes that they would like to achieve, and the DSS cannot work backwards to find the ideal inputs to obtain this result. On the other hand, inference engines utilise a process of backward chaining to assist decision making. Drawing upon an existing knowledge base of facts and rules, the inference engine works backwards from the desired outcomes to make predictions and suggest appropriate solutions. They can in fact take advantage of a computer's processing power to create, or infer, new facts and rules to solve the problem more effectively. Ultimately, inference engines test ideas to find the optimum solution to a given problem, without the user having to manually enter and test data by trial and error.

4. With regard to forward chaining and backward chaining:
a) Outline each of the processes.
Forward chaining starts with known facts and inputted rules, then works through these to reach a conclusion. This process relies on the user to manually input and test data; thus finding the most appropriate solution is largely through repeated trial-and-error. Backward chaining draws upon existing knowledge to work backwards from the desired outcomes and find the initial data that produces this optimum result. This process takes advantage of the computer's processing power in order to infer new rules and generate the ideal solution.
b) For each, identify two examples where it would be suitable where its counterpart would not be.
[do this later]
c) Based on the table of HSC electives, use both forward and backward chaining to come up with a list of alternative solutions for the subjects you will choose for next year, and thoroughly document the process.
[do this later]

5.4: Processing and decision support - Activities (pg. 206, Jacaranda)

Think
4. Explain the functions of the four major parts of an expert system: the knowledge base, the database, the inference engine, and the user interface.
The knowledge base is the main store of data, rules, known facts, and relationships that the expert system relies upon to make its decisions. The knowledge of various human experts is digitised and entered into this data store in a way that is accessible and usable by the expert system. This incudes rules used to solve problems; both straightforward textbook knowledge and chains of rules that depends on each other. The database is similarly a store of data, however it holds facts about the problem and the current status of the problem-solving process. It essentially contains everything that is known about the problem, and documents what rules have been used and in what order. The inference engine is essentially the brain of the expert system. It draws upon data in the knowledge base to find and suggest appropriate solutions, attempting to match facts with rules. Finally, the user interface is the means by which the software interacts and communicates with end users. It generates questions and possible answers for the user based on knowledge base rules, accepting this data and presenting the final output of the system as a solution to the problem.

5. Consider an expert system with certain pre-defined rules.
a) Use forward chaining to prove C. Write down the rules that you use and a list of inferred facts.
It is initially known that A, B, D, and R are true.
From rule 1, S is true (A, D both true).
From rule 2, P is true (B, R both true).
From rule 4, M is true (S, P both true).
From rule 5, C is true (B, M both true).
b) Did you use all the rules in part (a)? Were all of the initially known facts necessary to prove C?
Not all rules were used, however all of the initially known facts were necessary at some point.
c) Use backward chaining to prove C. Write down the rules that you use and a list of inferred facts.
We intend to prove that C is true. Let us assume that it is.
From rule 5, B and M must be true.
From rule 4, S and P must be true for M to be true.
From rule 2, B and R must be true for P to be true.
From rule 1, A and D must be true for S to be true.
Thus for C to be true, B and M must both be true (B is given as true).
Thus for M to be true, all of A, D, B, and R must be true.
However, this matches up with the initially known facts that A, B, D, and R are true.
Hence C must be true.

28 July 2009 (Tuesday)

posted Jul 27, 2009, 4:24 PM by Unknown user   [ updated Aug 4, 2009, 8:04 AM by Eddie Woo ]

Spreadsheet DSSs

Create a spreadsheet designed to measure the overall physical fitness of an individual and give them feedback on what areas they need to work on.
File is attached.

24 June 2009 (Wednesday)

posted Jun 24, 2009, 6:31 AM by Unknown user   [ updated Aug 3, 2009, 3:47 PM by Eddie Woo ]

Term 2 Exam Review

My overall mark was 19 + 65.5 = 84.5/100, which was surprisingly first in the class.

This exam was supposedly shorter in length than our previous exam yet we were given half an hour extra. As such, I deliberately worked at a slightly slower pace to give myself more time to think. As such, I finished with about 3 minutes to spare - just enough time to quickly check over the multiple choice section.

Multiple choice:
  • 6. I knew that statement II was definitely correct (high resolution displays are associated with a large frame buffer), however I assumed that statement I (high resolution displays are associated with smaller pixels) was incorrect because it is not always true - the question did not specify that the screen size remained constant. A higher resolution in an identically-sized panel would of course result in smaller pixels, but in many cases high-resolution displays are physically larger in size. Though there is some ambiguity in the question, I realise my mistake.
Question 21:
  • a) Did not know what AAC stood for, took a guess at "Advanced Audio Codec". Need to revise basic things like this before the exam.
  • b) Lost half a mark for not remembering what bit rate is. I understood it before, just forgot about it during the exam.
  • c) Full marks.
Question 22:
  • a) Identified it as compression, but not specifically as JPEG. Should have paid more attention to the "hint" Mr. Woo gave before the test.
  • b) to f) Full marks.
Question 23:
  • a) Did not read the question. Instead of describing the topology of just the library servers, I wrote a rambling explanation of the topology of the entire school network, which obviously cannot be simply described as one of the main five topologies. Need to remember to read and understand every part of the question.
  • b) to d) Full marks.
Question 24:
  • a) I found this question somewhat difficult.
    • Error detection and correction only occurs at the receiving end, not within the actual medium.
    • Data packets are arranged only once, at step 2.
    • The signal is only amplified within the medium, not at transmission/routing.
    • Handshaking does NOT occur at every stage; but only at the transmitter, switching/routing, and receiver.
    • The signal is only generated once by the transmitter, and merely passed on at switching/routing.
  • b) Correctly identified the process, however lost a mark in part ii) for not providing enough relevant detail (was missing identity theft).
Question 25:
  • a) Only got 7 marks out of a possible 10. While I identified plenty of examples of uses for multimedia and communication systems, I did not really critically analyse them - "identified relationships between the situation’s components, and inferred implications of the former". HSC keywords are important.
  • b) 5 marks out of 6. Perhaps needed to include more key points (e.g. space saving) and justify more benefits against the negatives of initial hardware and training costs.
Question 26:
  • a) Full marks.
  • b) One of the items i described under "multimedia" was more of a "communications" related technology. I was having problems coming up with other items under multimedia.
  • c) 3 marks out of 5. I missed out a few important points (forgot about them during exam conditions):
    • changing nature of work
    • unintended or malicious invasion of privacy
    • censorship vs. freedom of speech (hinted at later in the question)
    • reliability, bias, accuracy, integrity of data
    • breach of copyright
  • d) Full marks.

19 June 2009 (Friday)

posted Jun 18, 2009, 5:32 PM by Unknown user   [ updated Jun 18, 2009, 5:40 PM by Eddie Woo ]

Sample questions
 
Describe situations in which each of the major network topologies would be the most appropriate.
[do this later]

Converting a large organisation's information system from a wired infrastructure to a wireless infrastructure is a costly exercise. Identify specific costs involved and justify the conversion in spite of the cost.
[do this later]

The internet used to be called "cyberspace", which alluded to the fact that things and people on the internet inhabited another world and had no relation to the real world. This is simply no longer the case. Discuss this reality with reference to real events that have been changed or made possible only by the use of the internet.
[do this later]

1-10 of 54