Text Classification
NER, POS
Prashant K. Sharma
Diptesh Kanojia
https://shala2020.github.io/
We will start at 09.05 pm
1
Agenda
● Text Classification
● Classical approaches for text representation
● Modern approaches for text representation
● Challenges in NLP : NER, POS
2
Introduction to
Text Classification
NLP and Speech Layers
Credit: Prof. Pushpak Bhattacharyya 4
NLP and Speech Layers
● Morphology : karna / try - tried / visible - invisible / create - recreate
An example
● POS : ram.NN kicked.?? the.?? ball.??
● Chunking : the book has many chapters / noun compounds
● Parsing
● Semantics Extraction - Word Senses / Word Vectors
● Discourse - Coreference Resolution / Sarcasm Detection
Credit: Prof. Pushpak Bhattacharyya 5
Why do we need to classify Texts ?
● Spam Classification
6
Why do we need to classify Texts ?
● Sentiment Analysis
7
Why do we need to classify Texts ?
● Fake News/Click bait Detection
8
Why do we need to classify Texts ?
● Data Filtering
9
Text Classification in general
10
Text Classification in general
11
Text Label types
12
Text Classification in general
13
Classification Algorithms
14
Text Classification in general
15
Text Representation: Feature Engineering
16
Sparse Text Representation: BOW
17
Sparse Text Representation: BOW
18
Sparse Text Representation: BOW
19
Sparse Text Representation: BOW
20
Sparse Text Representation: BOW
21
Sparse Text Representation: BOW
22
Sparse Text Representation: BOW
23
Sparse Text Representation: BOW
24
Sparse Text Representation: BOW
25
Sparse Text Representation: BOW
26
Sparse Text Representation: BOW
27
Weighting techniques for BOW
28
Weighting techniques for BOW
29
Weighting techniques for BOW
30
Weighting techniques for BOW
31
Context Importance
32
Context Importance
33
Dense Text Representation: NBOW
34
Dense Text Representation: NBOW
35
Dense Text Representation: NBOW
36
Dense Text Representation: NBOW
37
Dense Text Representation: NBOW
38
Dense Text Representation: NBOW
39
Dense Text Representation: NBOW
40
Dense Text Representation: NBOW
41
Dense Text Representation: NBOW
42
Dense Text Representation: NBOW
43
BOW and NBOW: the shared problems
44
BOW and NBOW: the shared problems
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
References
● CS224n: Natural Language Processing with Deep Learning
● NLP course by Yandex data school
● https://medium.com/data-from-the-trenches/text-classification-the-first-step-to
ward-nlp-mastery-f5f95d525d73
68
Thank You!
69
The Hitchhiker's Guide to the Galaxy by Douglas Adams’s 70
The Hitchhiker's Guide to the Intelligence TM! 😃 71
Don’t Panic Guide to AI
What ? What ?
AI Concept How ? Cricket How ?
Why ? Why ?
72