Smarter Literature Review with NotebookLM (with Data Tables) — Transcript

Full transcript

1. 0:00Hello everyone. I'm pretty sure a lot of
2. 0:02you have been using Notula M for
3. 0:04different purposes, but in this video I
4. 0:07will be showing you how you can use
5. 0:08Notebook M to do better literature
6. 0:10review. Let's go.
7. 0:14[music]
8. 0:20So, as usual, once we go into Nobel, the
9. 0:23first thing we need to do is to create a
10. 0:24new notebook. Let's create a new one.
11. 0:27And you just drag all your sources in.
12. 0:29Imagine I have all these seven articles
13. 0:31about self uh regulated learning. So I'm
14. 0:33going to put a title here.
15. 0:34Self-regulated
16. 0:36learning. Okay. What happened is Noah M
17. 0:40will try to understand or summarize all
18. 0:43these sources that we have identified.
19. 0:45You can see everything is about
20. 0:46self-regulated learning here. Now in
21. 0:48literature view there are few things
22. 0:50that we can do with um Noel M. Uh if you
23. 0:53see the studio part here you have a few
24. 0:55things but uh we only need to focus on
25. 0:57two main uh studio tools. One is called
26. 1:00report. If you click report here
27. 1:03uh the default is briefing dog study
28. 1:06guide and blog post but in our case
29. 1:08because we want to do literature review
30. 1:10we can wait until uh no bulim run is uh
31. 1:14suggested format for us. We can go to
32. 1:16literature review here or if you don't
33. 1:19see this what you can do is click create
34. 1:22your own and then you put in your uh
35. 1:25instruction here for example uh
36. 1:28literature review of the articles right
37. 1:33focusing on you can kind of give your uh
38. 1:36instruction here or the prompt here
39. 1:38before you click generate I'm going to
40. 1:40use the one suggested here you can see
41. 1:42the suggested one synthesize the
42. 1:44provided academic papers into
43. 1:46comprehensive uh literature review
44. 1:48because this one is suggested by uh
45. 1:50Gemini on its own. You can see it's
46. 1:52related to self-regulated learning SRR
47. 1:55in Malaysian context. Okay, you can even
48. 1:57say adopt a formal academic tone to
49. 2:00provide scholarly overview. If you need
50. 2:02this, you can just wait for uh GBI to
51. 2:05suggest. So I'm going to click generate
52. 2:07now. [snorts] So by [clears throat] now
53. 2:10in the studio section, you will see the
54. 2:11report is being generated. So while
55. 2:13waiting for that to happen, I will show
56. 2:15you another one called data table. Why
57. 2:18do we need to use data table? You know
58. 2:20the typical thing that we normally do in
59. 2:22literature review or systematic
60. 2:24literature review is to have an overview
61. 2:26of what we have, right? And one of the
62. 2:28best ways to do this is to of course put
63. 2:31it in a table format or some people call
64. 2:33it the matrix, the table matrix, right?
65. 2:35by compartmentalizing
66. 2:37all these uh key areas or key components
67. 2:40of the literature of the or the
68. 2:42articles. Right? So what you can do is
69. 2:44click this one but don't click the data
70. 2:48uh data table immediately. Click these
71. 2:50uh the p pencil icon or the pen. Click
72. 2:53on it and then you will see this uh
73. 2:55language and of course you need to
74. 2:57describe the table that you want. So I
75. 2:59will put a make a table
76. 3:03uh reviewing these articles
77. 3:06according to this column. So you can say
78. 3:11maybe you need authors title, maybe you
79. 3:15also need to know the method,
80. 3:18right? Maybe you need to know the key
81. 3:20findings or you can also put
82. 3:21implications and lastly maybe research
83. 3:26uh gap. All right. So, just put the
84. 3:28columns that you need here. All these
85. 3:31columns will kind of help you to
86. 3:34strategize the um the table format. Just
87. 3:37click generate.
88. 3:39Wait for it to happen. Uh usually it's
89. 3:42very fast. If you click the data table
90. 3:44directly, what happened was uh Gemini or
91. 3:47no m will decide for you. All right?
92. 3:49We'll decide on your behalf. I think
93. 3:51sometimes better for us to set the
94. 3:53structure right so that we don't um you
95. 3:56know get things that we don't want.
96. 3:57Okay. All right. See the table is ready
97. 4:00now this one. So let me click you will
98. 4:02see the table now right. Wonderful. All
99. 4:05right. All this table is sorted
100. 4:07correctly now. You can see authors and
101. 4:10all this and then source ones and you
102. 4:13know resource two. Amazing right? And if
103. 4:17you like to verify, of course, you can
104. 4:19always uh click all this and then verify
105. 4:22uh whether it's true. But I have done a
106. 4:24few of checking uh you know previously
107. 4:27uh for the last few days and I noticed
108. 4:29is really accurate in extracting the
109. 4:31needed parts. What is probably causing
110. 4:34some problem will be if the article did
111. 4:37not mention any of this. For example,
112. 4:38the article did not mention about
113. 4:40research gap. um it will try to decide
114. 4:43based on the way the research problem is
115. 4:46worded. I think it's still fine because
116. 4:48we get um we kind of get the you know
117. 4:51bird's eye view of what's happening and
118. 4:53it's easier for us to decide which
119. 4:55article should we go deeper right so
120. 4:58this is really really amazing of course
121. 5:01if you need to uh show the prompt you
122. 5:03can get the prompt again right let me go
123. 5:06back to studio here if I were to do
124. 5:09another table I will probably click data
125. 5:11table again I will paste the prompt that
126. 5:14I have just now and then I will add
127. 5:16maybe a reference
128. 5:20in APA7 edition. Let's try this, right?
129. 5:23Let's try this. It's not really 100%
130. 5:25accurate all the time, but when you do
131. 5:27this, at least you also get the
132. 5:28referencing. All right, so let's
133. 5:31generate.
134. 5:32While that is being generated, you can
135. 5:34see the literature review is already
136. 5:36done here, the report just now, and it's
137. 5:39done in such a way. Okay? Right. So you
138. 5:43can also see the uh the review done for
139. 5:46you. Now I wouldn't say is that
140. 5:49extensive yet but you get the skeleton,
141. 5:51you get the draft and you can quickly
142. 5:54you know expand based on your own
143. 5:56understanding. Of course we don't have
144. 5:57to rely fully on AI. This is just a
145. 6:00draft for us to shape our thinking in
146. 6:02terms of what we can do with our
147. 6:04literature review. So I wouldn't suggest
148. 6:06you to just copy blindly. In fact, you
149. 6:08should not copy blindly. What you need
150. 6:10to do is perhaps see the structure and
151. 6:12any part that you feel is lacking, you
152. 6:14can prompt further and then you probably
153. 6:16add your own words inside. All right. So
154. 6:18that is one way. Now even though now you
155. 6:21have this function of export to doc let
156. 6:24me try to export this to let you see it
157. 6:27will export to document and this is how
158. 6:29it looks like. Not really clean at the
159. 6:31moment but it's good enough to have a
160. 6:33Google doc version and then we can do
161. 6:34the formatting uh later. All right. Um,
162. 6:38of course, let's get back to the new
163. 6:40table that we asked just now. You can
164. 6:42see that the table is also including the
165. 6:46APA uh referencing. But unfortunately,
166. 6:49you will see how to put it. Uh there are
167. 6:51some error here where you know all the
168. 6:53sources seems to be unable to go through
169. 6:56everything. All right. Partly because it
170. 6:59could not extract some of the references
171. 7:01here. So this is something that you have
172. 7:03to be careful of when you request
173. 7:05something from the paper and the paper
174. 7:07doesn't seem to have adequate uh or
175. 7:09enough information. It will you know it
176. 7:11will not be able to provide accurate
177. 7:13ones. If you compare this one just now
178. 7:15with the one that I had earlier it's
179. 7:17accurate because everything I asked for
180. 7:19is there. Okay, that's something for you
181. 7:21to pay attention to. But if you would
182. 7:23like to export this, I kind of recommend
183. 7:26you to also export it as table uh sorry
184. 7:29uh sheets right to Google Sheets. If I
185. 7:31do this, it will go to Google Sheets. So
186. 7:33you can still have a good uh compilation
187. 7:36of what you need in this format. Okay,
188. 7:39so that's about all for how you can
189. 7:41actually do better uh literature review
190. 7:44using Notebook LM. Of course, there are
191. 7:46other prompting techniques that we can
192. 7:47do, but I'm just showing you the key
193. 7:50features that you can maximize your
194. 7:52literature review process. Uh, if you
195. 7:55want to synthesize more and all that,
196. 7:56that will require another prompting
197. 7:58strategy. But in this video, the focus
198. 8:01is more on the existing data table and
199. 8:04also report features that you can use
200. 8:06for your literature review. Happy
201. 8:08trying.
202. 8:17>> [music]