Ask4Summary Documentation

Here is the Documentation for Ask4Summary. This content is also found in the block itself.

Terms of Use

The VIP Research Group is a research group led by Prof. Maiga Chang (https://www.athabascau.ca/science-and-technology/our-people/maiga-chang.html) at the School of Computing and Information Systems, Athabasca University. This plugin, Ask4Summary, is one of the research group's works. The research group does have a follow-up research plan to improve it and further use it in other research projects.

Almost all of Prof. Chang's works are open access (or open source). This plugin is now open access and there is no plan to make it open source. The N-Gram POS service used by the plugin is open access and running on a self-sponsored server, as all of other research projects (see http://maiga.athabascau.ca/#advanced) they will be always online, improving, and accessible as long as the cost can be affordable and covered by Prof. Chang.

Of course if in any case just like the access volume of the web service becoming high or any business/commercial takes advantage of using it to make money, then the term of using the web service may look for changes; for examples, donations, personal/academic/business license and subscription modes, etc. However, it is really too early to say that.

What is Ask4Summary?

Ask4Summary is a plugin designed to use the VIP Research Summary Generation algorithm to produce a response to a user query relating to their course. This is implemented in Moodle by utilizing the forum feature which allows course users to post a question and have the plugin generate a response.

This forum can be designated by a teacher and/or an administrator and will check for posts which contain the set "Helper Name" - what users will refer to as Ask4Summary in their question. It will then use the Summary Generation algorithm to produce a response, and directly respond to the same forum posting.

The Ask4Summary plugin also has an additional function, which is the ability to scan and parse course modules text content. These course modules (files such as PDFs, Word Documents, URLs and Powerpoint Presentations) will be then used as a learning tool for the plugin.

How does the algorithm work?

The VIP Research algorithm works by using Part of Speech analysis, a method which breaks down words by what type they are (noun, adjective, verb, etc). This is done by taking the user question and breaking it apart, sentence by sentence. Afterwards, it analyzes the most important words of that sentence, called "N-Grams", and retrieves their Parts of Speech. Then, with use of "Cosine Similarity", it will compare the N-Grams taken from the user question to the frequency of N-Grams taken from the course content. Then, by comparison to every resource, it will take the sentences that relate the best to the user question.

However, there are some limitations of the algorithm. Since the answer is based on frequency of the top Part of Speech N-Grams, certain words not may not be recognized. These words may be critical to the user question, and may result in a different result entirely. So our recommendation is to stick to singular nouns and adjectives in your query.

For more information, see the following webpages:

N-Gram POS Service
AskCOVIDQ Service
What are N-Grams and Parts of Speech?
What is Cosine Similarity?

Where are the settings changed?

All of the plugins settings are controlled in block configuration. To arrive there, simply turn editing on for the course, click the gear on the top right-hand side of the plugin, and choose "Configure Ask4Summary Block."

Once on the edit form screen, there will be several options to choose from:

  • The Ask4Summary Helper Name
    • This is what students need to include within their forum posting (either in the subject or the question itself) to trigger Ask4Summary to answer their question
  • Whether the response functionality should be enabled
    • The checkbox may or may not be greyed out, depending if an administrator has disabled the ability to change it
  • The forum response type
    • Consider every single forum within the course (this is the default)
    • Consider only a specific forum that already exists in the course (this can be chosen)
    • Create a forum in the course, and only consider that (the name can be selected)

It is important to note that these settings will only affect the course which the block was accessed. So, if you have the plugin put in multiple courses, and would like to change the administrator defaults, you will need to individually change each course blocks settings.

There are also course module scanning options to choose from, which will be explained in the following section.

What are course modules, and what does Ask4Summary use them for?

Course modules are activities and resources that are available within the course. Good examples of course modules are things such as assignments, forums, files, and webpage links.

Ask4Summary scans and parses course modules for their text content to build its N-Gram and Part of Speech database, for learning algorithm purposes.

The recommendation for improving the quality of sentences parsed would be to include resources and files which are strictly text based. Text based in images cannot be parsed, and documents with special formatting may be parsed differently than expected.

The course modules that Ask4Summary currently supports scanning for are:

  • URL webpages
  • PDF files
  • Word Documents
  • Powerpoint Presentations
  • Moodle Page Modules

There are additonal settings also found in "Configure Ask4Summary Block" that specifically relate to course module scanning and parsing.

These settings include:

  • Whether webpage scanning and parsing should be enabled
  • How many webpages found within webpages should be considered when parsing
  • Whether PDF file scanning and parsing should be enabled
  • Whether Word document scanning and parsing should be enabled
  • Whether Powerpoint scanning and parsing should be enabled
  • Whether Moodle Page scanning and parsing should be enabled

PDF parsing is disabled by default. Please see the PDF: AbiWord section for more information on why.

PDF: AbiWord

Our personal recommendation for the best course modules to use would be Word documents, Moodle pages, and PDFs, generally in paragraph format. In a format like this, it makes it very easy for the plugin to gather text content.

The other two formats, URLs and Powerpoints, are variable in format which can be tricky for the plugin to gather text content. If you would like to use these, consider keeping the crawling depth to 1, including webpages with paragraph content, and to have the Powerpoint slides in some kind of sentence format.

How is a question asked? How long does it usually take?

Questions are asked by going to the selected forum option, and including "Hi (Helpername)" in the forum post, or including exactly that in the subject of the forum post.

For example, if you would like to ask "How long should my report be?" and your helpername is set to "A4S", you would write your forum post as "Hi A4S How long should my report be?", or have your subject as "Hi A4S", then simply have the post contain the question.

The time for a question to be answered depends entirely on the length of the question and the amount of files parsed at the time of the question. Generally, however, the process is relatively quick.

Can I control how often questions or course modules are scanned?

If you are an administrator, you can control the intervals for forum scanning and document scanning from the "Site Administration: Server" tab.

If you are not an administrator, the default intervals are unobtrusive and will not slow down course performance. Course modules, which will certainly take the longest to scan and parse, happen from midnight to six in the morning, so class performance will not be reduced. Forum post scanning is minor, and once posts are scanned it will not consider them again.

How can I see the progress of Ask4Summary processes?

To see the progress of Ask4Summary processes, head over to the "Logistics" tab, either from the top of this webpage, or from the block itself in your course. Once here, you will be able to see the relative progress of forum posts scanned, forum posts answered, course modules scanned, and course modules remaining.

PDF: AbiWord

PDF Processing with Ask4Summary is disabled by default because it requires the external application "AbiWord".

AbiWord is a free, open source software that is used for word processing. It provides a wide variety of functionality; but, Ask4Summary uses it to get the text content from a course module PDF.

Ask4Summary will still be functional without the use of this application; however, PDF parsing will not be able to function correctly. Since Ask4Summary needs AbiWord to get the text content, without it, there will be no text content generated from a PDF file. This will record the PDF file in the Ask4Summary database with no text content, rendering the course module useless.

If you are unsure if your administrator has installed AbiWord, contact them, or simply keep PDF parsing disabled.

For administrators, to install AbiWord on Ubuntu:

  1. Access the terminal
  2. Enter "sudo apt install abiword"
  3. Enter your password

Or if you would prefer to use the Software Center Ubuntu has:

  1. Open Ubuntu Software Center
  2. Search for AbiWord
  3. Go to the install page for AbiWord
  4. Click install
  5. Enter your password

For other distributions of Linux, please see the AbiWord site, or the additonal link provided:

AbiWord
Other Linux Distributions