1.
3 Strategies
for
Business
Innovation
with
Sentiment
Analysis (ML)
Rahul Ramachandran, PMP®, MBA
Amancio Bouza, PhD, ipt, Innovation Process Technology AG
Cloud Days 2018, Luzern

2.
2
The Challenge..

3.
There has been a paradigm shift across industries
19601900 1990 2010
The age of..
MANUFACTURING
INDUSTRIAL REVOLUTION
What triggered it..
The paradigm..
OWN PRODUCTION,
OWN THE MARKET
DISTRIBUTION
GLOBALIZATION
& FREE TRADE
OWN DISTRIBUTION,
OWN THE MARKET
INFORMATION
INTERNET
OWN INFORMATION,
OWN THE MARKET

4.
This is no rocket science – Happy customers return and refer..

5.
..and what are today’s customers expecting
..but speed alone doesn’t cut itSpeedy contact and response

6.
So we want to respond, but..
“Customer Experience”
is EVERYTHING
What is the biggest challenge for
businesses?

7.
So we want to respond, but..
“Customer Experience”
is EVERYTHING
Too many inputs

8.
..and too many platforms to read from
Too many feedback
platforms
Need to
leverage
“Customer Experience”
is EVERYTHING
BUT
HOW?

9.
The Opportunity..

10.
In every challenge lies an OPPORTUNITY
Customer Retention
Product Development
Competition Analysis
Business Potential

11.
To harness the potential – BIG Data & Machine Learning (ML)
• Machine learning is the subfield
of computer science that gives
computers the ability to learn
without being explicitly
programmed (Arthur Samuel,
1959). ...
• Machine learning is closely
related to (and often overlaps
with) computational statistics,
which also focuses in prediction-
making through the use of
computers.

12.
Machines are now able to process natural human language
and also derive the human sentiment within
12 Source: Stanford NLP

13.
Machines are now able to process natural human language
and also derive the human sentiment within
13

14.
Natural Language Processing (NLP) is very a HOT topic

15.
So how did we approach it..

16.
So how did we approach it..

17.
..using Microsoft’s Azure ML and Cognitive Services

18.
18
Wrapping up – What we achieved..

19.
A detailed analysis of our customer feedback
19
Sentiment: Positive
Entity: Michael (Person)
Entity: Luzern (Location)
Input: “Micheal was a great guide and we enjoyed Luzern city very much”

20.
In our initial POC, we received a high accuracy
20
79%accuracy 75% precision 73% recall

21.
“Word cloud” on which location are being talked about
21
Enables business to:
- Quickly focus on the “most”
talked locations
- Setup effective communication to
resolve most issues
- React FAST
- Incorporate feedback to Product
Development
- Understand what your market
peers are doing right

22.
Contact
Rahul Ramachandran
Director Business Solutions
Viking Cruises
rahul.ramachandran@vikingcruises.com
22
Amancio Bouza
Principal Consultant
ipt, Innovation Process Technology AG
amancio.bouza@ipt.ch

23.
famous for integration