Whats-For-Dinner-Data-Science-Project/methods.html at main · LegoMasterBuilder/Whats-For-Dinner-Data-Science-Project · GitHub

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
<!DOCTYPE HTML>
<!--
	Miniport by HTML5 UP
	html5up.net | @ajlkn
	Free for personal and commercial use under the CCA 3.0 license (html5up.net/license)
-->
<html>
	<head>
		<title>PH Twitter Fake News Analysis</title>
		<meta charset="utf-8" />
		<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
		<link rel="stylesheet" href="assets/css/main.css" />
		<style>
			pre code {
				background-color: #eee;
				border: 1px solid #000;
				display: block;
				padding: 1px;
				text-align: left;
			}
		</style>
	</head>
	<body class="is-preload">
		<!-- Nav -->
			<nav id="nav">
				<ul class="container">
					<li><a href="index.html">Home</a></li>
					<li><a href="about.html">About</a></li>
					<li><a href="datapreprocess.html">Data</a></li>
					<li><a href="methods.html">Methods</a></li>
					<li><a href="visualization.html">Visualization</a></li>
					<li><a href="results.html">Results</a></li>
					<li><a href="conclusions.html">Conclusions</a></li>
				</ul>
			</nav>

		<!-- Home -->

		<!-- Conclusion and Implications -->
		<article id="results" class="wrapper style3">
			<div class="container">
				<header>
					<h2>How was the research conducted?</h2>
				</header>

				<p>Our research followed a standard data science methodology. We began by collecting Twitter data from 2016 to 2022
					using Twitter APIs. The extracted raw data was stored in a Google Sheet document, including information such as bio,
					account name, account type, location, date created, and time posted. We then proceeded with NLP preprocessing techniques,
					which involved removing irrelevant information like URLs, usernames, and special characters. Additionally, we lowercased
					the text, performed lemmatization, and eliminated stopwords. To ensure consistency, we translated Filipino tweets to English,
					it is also here where we fixed wrong spelling of terms.</p>

                <header>
					<h2>What is the data about?</h2>
				</header>
                <p>The dataset used in our research consists of Twitter data spanning from 2016 to 2022.
					It includes a variety of information associated with each tweet, such as user bio, account name, account type,
					location, date created, and time posted. The focus of the research was to analyze tweets related to the food
					situation during the Ferdinand Marcos regime, specifically targeting potential misinformation about it.</p>

				<header>
					<h2>What methods did we use to collect and analyze the data?</h2>
				</header>
				<p>To collect the data, we employed Twitter APIs, which allowed us to access and extract relevant information from Twitter.
					The extracted data was then organized and stored in a Google Sheet document for further analysis. For data analysis,
					we utilized several methods. Firstly, we performed NLP preprocessing steps to clean the collected tweet data, including
					removing irrelevant information, lowercasing, lemmatization, and stopword removal. Additionally, we took the necessary
					step of translating Filipino tweets into English to maintain consistency. After preprocessing, we applied the TF-IDF
					vectorization technique to convert the preprocessed tweet data into numerical feature vectors. Subsequently, we employed
					Latent Dirichlet Allocation (LDA) as a topic modeling technique to identify latent topics within the tweet data. To visualize
					the resulting clusters, we employed t-SNE for dimensionality reduction, reducing the high-dimensional feature vectors to
					two dimensions. Finally, we used Plotly Express, a data visualization library, to create interactive visualizations that
					effectively showcased the identified clusters in the reduced 2D space.</p>
			</div>
		</article>

		<!-- Team -->
		<article id="team" class="wrapper style4">
			<footer>
				<ul id="copyright">
					<li>&copy; Nutribuns. All rights reserved.</li><li>Design: <a href="http://html5up.net">HTML5 UP</a></li>
				</ul>
			</footer>
		</article>


	</body>
</html>