-
Notifications
You must be signed in to change notification settings - Fork 3
Expand file tree
/
Copy pathnotes-en.html
More file actions
300 lines (207 loc) · 19.3 KB
/
notes-en.html
File metadata and controls
300 lines (207 loc) · 19.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, maximum-scale=1">
<title>Girls Learning Code</title>
<link rel="stylesheet" href="framework/css/slideshow.css" data-noprefix>
<link rel="stylesheet" href="framework/css/fonts.css" data-noprefix>
<link rel="stylesheet" href="framework/css/highlightjs/github.css" data-noprefix>
<link rel="stylesheet" href="framework/css/styles.css" data-noprefix>
<link rel="shortcut icon" href="framework/img/favicon.ico">
<!-- Takes care of CSS3 prefixes -->
<script src="framework/scripts/prefixfree.min.js"></script>
<!-- opens all links in a new window -->
<base target="_blank">
</head>
<!-- Timer/progress bar: Define the presentation duration using "data-duration" in minutes. -->
<body class="en" data-duration="360">
<main>
<section class="slide" data-toc data-markdown>
<script type="text/template">
##Instructor Notes
##Girls Learning Code: Tackling Cyberbullying with Machine Learning
<br>
#About the Content
<div style="font-size:25px;">
Cyberbullying is a complicated problem which requires an out of box solution. Learners will get the opportunity to learn about the harmful effects of cyberbullying and methods for combating it.
Learners will explore the limitations of traditional programming in solving complex issues. Learners will learn about Machine Learning and how it differs to the programming they are used to through unplugged activities. The final portion of the workshop learners will train and customize an intelligent assistant!
Machine Learning can seem like a daunting topic, however this content is scaffolded and topics are introduced with unplugged activities to make it digestible and fun. If you would like more information about machine learning, <a href="https://towardsdatascience.com/explaining-supervised-learning-to-a-kid-c2236f423e0f">this article provides a good explanation</a>.
I hope you enjoy teaching this content as much as I enjoyed making it! Best of luck, you got this!
<br>
#Key Programming Concepts
<ul>
<li> <strong>Conditional Statements</strong> use a condition to decide whether an action is executed. If the condition is evaluated to be true, some operation is performed.</li>
<li> <strong>Loops</strong> are a computer action that is repeated until a specified condition is reached.</li>
<li> <strong>Data</strong> is distinct pieces of information, usually formatted in a special way. </li>
<li> <strong>Machine Learning Algorithms</strong> are algorithms with the ability to automatically learn and improve from experience without being explicitly programmed.</li>
<li> <strong>Supervised Learning</strong> is a training process which takes a known set of data (inputs) and its known responses to the data (outputs) to learn how to perform a task. </li>
</ul>
#Schedule
 What is Cyberbullying (Unplugged) (50 min)
 Introduction to Machine Learning (40 min)
 Training our Intelligent Assistants (30 min)
 Code-Along in Scratch (40 minutes)
 Demos (20 min)
<br>
#Materials + Special Requirements
* This activity requires WiFi
* This activity requires hard copies of "The Social Butterfly Effect"
* This activity requires pencils and/or markers
* This activity requires post-it notes
* The activity requires paper
#Contingency
* In the case of no WiFi the workshop's first 2 sections remain unchanged
* When the time comes to train the assistant, learners should use <a href="https://scratch.mit.edu/download">offline scratch</a> to create an assistant which looks for keywords
* If there is time remaining, perform activity 2 from <a href="https://www.canadalearningcode.ca/lessons/learn-like-a-computer/">Learn Like a Computer</a>
#Mentor Huddle Talking Points
* Mentor Names
* Overview of the workshop & schedule
* Disclosure - if any learners want to talk about their experiences, direct them to (kids help phone resource); if they share that they or someone else could be in immediate danger (life is threatened) alert (chapter lead).
* Instructions for mentors: Facilitating conversations with learners and assigning 1 mentor per table for the first half of the workshop
* Ask mentors if they have an questions
Once the mentor huddle is complete, ask all mentors to help learners interact with each other. Ask them to use the slide 1 questions as a conversation starter.
<br>
#Introduction
**Introductions (Slide 3):**
<br>Ask learners to put their computers into "politeness mode" (closed halfway) before you begin.
Have mentors stand up and introduce themselves.
**Icebreaker (Slide 4):**
<br>Choose an icebreaker from the list of <a href="https://docs.google.com/document/d/18zNvgQ9bngEjhWDkbtaA8Pw5v0JASYpMag4xi8GazJQ/edit">icebreaker activities</a>. Try to pick a game which promotes verbal interaction within learners.
**Coders' Code (Slide 6):**
<br>Ask: What are some ways you can ask for help during the workshop? (A: Peers, Mentors and Instructor)
Ask: What does it mean to be present? (A: Paying attention and following along with the activities). Use this as an opportunity to establish call back words such as “hands on top that means stop”and/or “waterfall”.
Ask: What does respect others mean? (A: Treat others the way they would like to be treated).
Ask: What does trying your best mean? (A: Even the best coders have trouble sometimes, so we don’t expect you to do everything perfectly the first time, but make sure you’re trying your best. Don’t be afraid to try things out and make mistakes!)
Ask: Is anything missing? What else do we need to have an awesome experience today?
**Disclosure (Slide 7):**
<br>Read off the slide. Canada Learning Code would like no paraphrasing of this statement.
Mention to learners that Kids Help Phone live chat uses a form of artificial intelligence to help you while you wait to connect with a real person. Today we will be learning about how that works!
Mention to learners that they will be receiving the contact information for Kids Help Phone to take home.
<br>
<br>
#What is Cyberbullying?
**Cyberbullying is... (Slide 13):**
<br>Ask: Are children and teens the only ones targeted through cyberbullying? (A: No! It can happen to adults as well ie) Hurtful emails or work group chats).
**No One Definition... (Slide 15):**
<br>Let’s define cyberbullying together! Each group should discuss with each other what they believe cyberbullying is. Each group with the help of mentors should write a definition of cyberbullying.
**Social Media (Slide 16):**
<br>Ask: Do you think social media is inherently bad? (A: No)
Ask: How can social media be positive? (A:Being able to stay connected with friends and family whom you can’t see all the time. Other examples can be an outlet to express yourself and bring people together.)
**Throw it in the Bucket (Stretch Goal):**
<br>Throw it in the bucket. Each learner should write a Social Media Platform they are familiar with on a piece of paper or post-it note.
After all have been collected each group should be given one piece of paper from the bucket with a unique social media platform on it.
Each group needs to act out the social media platform while learners guess which they are. Try to ensure at least one group is acting out <strong>Instagram, Snapchat and YouTube </strong>.
**The Social Butterfly Effect (Slide 17):**
<br>Have one learner from each group come up and get comic books for themselves and their peers. Note adults attended do <strong>NOT</strong> get a comic book.
**(Slide 18):**
<br>Learners should read the comic book together with group mates. A method of doing this is assigning each member of the group a role. The roles are:
Amira (Lots of reading)
<br>Social Media (Lots of reading)
<br>Rochelle (Moderate reading)
<br>Amira’s Mom (Little reading)
<br>Locker Girls (Little reading)
If any members of the group have trouble reading in English or French encourage them to read with a mentor or parent.
**Who were the bullies in Amira's story? (Slide 20): **
<br> (A: Rochelle did <strong>NOT</strong> mean to cyberbully Amira.)
The butterfly effect is a theory in math where the smallest action can have large effects. The name comes from the idea that the wind generated by the flap of a butterfly's wings can cause a tornado thousands of kilometers away.
**Who were the bullies in Amira's story? (Slide 21):**
<br>(A: The creators of the meme/video were bullies because they were mistreating/causing harm to Amira at the time when she was perceived as vulnerable.)
(A:Anyone who liked or shared the video were encouraging/approving/spreading the bullies behaviour, thus making them bullies as well.)
**THINK (Slide 23):**
<br>While going through the THINK acronym write what each letter means in a space where all learners can see. This definition will be referred to at a later time in the workshop.
Ask: What is a statement which would pass the THINK test?
**Who were the bystanders in Amira's story? (Slide 26):**
<br>(A:The locker girls along with anyone viewing the video/meme were bystanders because although they may not support the bullies views they were not doing anything to help Amira’s situation.)
**The Amazing Relay... (Slide 28):**
<br>The Amazing Relay should encourage learners to self explore strategies to help Amira from both her and a bystanders perspective. Encourage learners to stick their post-it notes in a common place such as a wall or window.
<br>
<br>
#Introduction to Machine Learning
**Code is... (Slide 33):**
<br>To demonstrate how specific programming needs to be have learners give you instructions on how to brush your teeth, make toast or make your bed. Act every instruction literally. For example, if a learner says to put the toothbrush in your mouth, pretend to put it on your tongue. The goal is for learners to give instructions such as “put the toothbrush horizontally on your teeth and then move your hand sideways.”
**Consider the Following... (Slide 35):**
<br>Ask a learner to say the statement on the screen out loud reflecting the tone of the message.
**Sentiments (Slide 39):**
<br>Creating cyberbullying filters are so difficult, because in text small changes such as an exclamation mark can change the meaning of a sentence. This is because it is hard to convey <strong>tone</strong> over text.
**Let's Collect Data (Slide 46):**
<br>(A: The colours purple, green, yellow, orange, blue, brown, beige and red are in the cart.)
(A: There are 3 fruits in the cart)
(A: There are 3 inedible things in the cart (shoe, hat and shirt))
**Let's Go on a Picnic (Slide 47):**
<br> Arrange students into a circle, about 5 people each. Choose 1 mentor to be the picnic organizer/trainer. Have this mentor choose a rule that all objects being brought on the picnic must follow. An example rule could be that the object must be yellow. Make sure no one else in the circle knows what the rule is!
The trainer should return to the circle and start by saying 'I am bringing -------- on my picnic.' The blank should be filled by an object that follows the rule they created. Go around the circle, one learner at a time. Each student says “Can I bring _____ on your picnic?”. The picnic organizer should respond with “yes” or “no”. The game should continue until all students have discovered the rule. Remind students who have discovered the rule that they shouldn’t reveal it. Once everyone has figured out the rule and if time permits allow learners to be the picnic organizer.
**Reflection (Slide 48):**
<br>
Let’s talk a little bit about learning!
Ask: What are some ways you like to learn?
Computers also have a favourite method of learning too: it’s trial and error! Computers love to learn from their mistakes!
Draw a number on a board or piece of paper and show students. Ask: What number is this? Follow up by asking how do you know that? (A: Because you have seen this number a thousand times).
Ask: Who has heard the term machine learning or artificial intelligence?
Ask: What is machine learning? (A: See terminology)
The process of learning based on data and human feedback is called supervised learning.
Ask: How did you make your first guess? (A: It was random!) Machine learning algorithms do the same thing, their initial guesses are completely random. This is because they don’t have enough information to make an educated guess.
Ask: What patterns were you looking for when deciding what to bring on the picnic? (A: Noticing a trend in the items that were allowed on the picnic). It was almost like connecting the dots, after you made enough connections in your head you could figure out the theme. This is exactly how machine learning algorithms such as a neural network learns.
Supervised learning is the process of an algorithm learning by seeing many examples (data). The algorithm then looks for patterns in the data to learn about what it is seeing.
<a href="https://www.canadalearningcode.ca/lessons/learn-like-a-computer/"> Full lesson plan: Learn Like a Computer</a>
**Muffin or Chihuahua (Slide 50):**
<br> What a mess! Let's see if there is a way that we can separate the chihuahuas from the muffins.
**Muffin or Chihuahua (Slide 51):**
<br>We need to look for ways to classify the images. These need to be measurable attributes about each image. In this example I have chosen the color of the muffin or chihuahua and the number of dark or black spots.
Choosing a diverse set of images will make the model more accurate. This means choosing 50% chihuahua photos and 50% muffins. Furthermore, choosing images with a variety of lighting and backgrounds! The more diverse the set of photos the better the model.
Notice now our muffins and chihuahuas are all grouped together!
**Muffin or Chihuahua (Slide 52):**
<br>This is the step which makes machine learning so powerful. Since we are training we know which image is a muffin and which is a chihuahua and our goal is to have our computer be able to distinguish the two.
We can draw a line that separates the muffins from the chihuahua’s! Machine Learning is all about picking the best line to separate the two. The math involved (Gradient Descent or Linear Least of Squares) in calculating this is well outside the scope of the workshop but there are many online resources for learning about it.
**Muffin or Chihuahua (Slide 53):**
<br>We have now successfully trained our computer on learning to distinguish between a chihuahua and muffins. Everytime we ask the computer “Am I looking at a chihuahua or a muffin” the computer graphs the image and checks which side of the line it is on! In 5 steps you now know the process for how most machine learning models are trained!
**Let's Gather Some Data (Slide 56): **
<br>A blank T chart is provided at the end of the comic book. Encourage learners to think of their own nice phrases. However, stress they are to only write mean phrases found in the comic book!
<br>
#Training our Intelligent Assistants Code-Along
* If there is no WiFi skip this code-along
* <a href="https://docs.google.com/document/d/1LYqT2h3PmMnqNW3qvmgWcyFXoEHEb3r6N4N-cVnlGAo/edit?usp=sharing">Solution Key is available here (Pages 1-4)</a>
**Over & Underfitting (Slide 58): **
<br>Explain these slides before learners begin to write their phrases into their labels in Machine Learning for Kids.
**Break (Slide 60): **
<br>Training the models usually takes around 5 minutes. Use this break as an opportunity to allow learners to relax while the models are training. Make sure learners do not exit the website or close their computers.
<br>
#Programming our Intelligent Assistants Code-Along
* <a href="https://docs.google.com/document/d/1LYqT2h3PmMnqNW3qvmgWcyFXoEHEb3r6N4N-cVnlGAo/edit?usp=sharing">Solution Key is available here (Pages 4-11)</a>
* <a href="https://docs.google.com/document/d/1WZjedcWpamACVFDS9WeKNXTK_afPmhr1pSdjTzUfiEg/edit?usp=sharing">Solution Key if there is <strong> NO WiFi </strong> is available here</a>. Ensure to notify learners that this is <strong> NOT </strong> machine learning. Explain that using keywords isn't a fool proof method. There exists nice and mean phrases not containing any keywords which will confuse the assistant. Furthermore, some keywords can be both nice and mean depending on context! This is why machine learning is needed! If there is time remaining, perform activity 2 from <a href="https://www.canadalearningcode.ca/lessons/learn-like-a-computer/">Learn Like a Computer</a>.
**Bias is... (Slide 62): **
<br>An example of bias is someone liking cats and disliking dogs having never interacted or owned a dog. This is a bias towards cats, because their opinion on dogs has no basis.
**Which Dataset is Better? (Slide 65)**
<br>The second dataset is the more diverse dataset. The first dataset has a selection bias towards shoes mainly worn by men. If a computer was trained using the first dataset, the computer could have trouble recognizing heeled shoes because it wasn’t exposed to enough examples of them.
<br>
# Gallery Walk (Stretch Goal)
* Each group of learners should have a place to display their assistant
* Encourage learners to have a 20 second elevator pitch about there presentation
* Remind learners to speak clearly and at a conversation pace
* After slide 72 give learners a few minutes to prepare for demos
* Encourage every learner to speak, even if its one sentence
</script>
</section>
</main><!-- cls main section -->
<script src="framework/scripts/jquery-1.11.0.min.js"></script>
<script src="framework/scripts/slideshow.js"></script>
<!-- Uncomment the plugins you need -->
<script src="framework/scripts/plugins/css-edit.js"></script>
<script src="framework/scripts/plugins/css-snippets.js"></script>
<script src="framework/scripts/plugins/css-controls.js"></script>
<!-- <script src="plugins/code-highlight.js"></script>-->
<script src="framework/scripts/plugins/markdown/marked.js"></script>
<script src="framework/scripts/plugins/markdown/markdown.js"></script>
<script src="framework/scripts/plugins/highlight/highlight-8.4.min.js"></script>
<script>hljs.initHighlightingOnLoad();</script>
<script src="framework/scripts/llc.js"></script>
<script>
var slideshow = new SlideShow();
// Grabs all the .snippet elements
var snippets = document.querySelectorAll('.snippet');
for(var i=0; i<snippets.length; i++) {
new CSSSnippet(snippets[i]);
}
</script>
</body>
</html>