Chapter 7 Flashcards
Customer satisfaction
Survey based. Offline.
qdap (Quantitative Discourse Analysis Package)
An R package designed to assist in quantitative discourse analysis. The package stands at a bridge between qualitative transcripts of dialogue and statistical analysis & visualization.
Connected to a sentiment analysis.
How can it be analyzed if results from an analysis are positive, negative, or neutral?
Using qdap R-package.
External unstructured data
Social media data
Blogs
Review sites
External structured data
Public data
ZIP code data
Household data
Internal unstructured data
Customer contact data (e-mail, text message, shop, website)
Mobile data
Internal structured data
CRM data
Sales data
Invoice data
Transaction data
Value
Usability.
Sits at the top of the big data pyramid. It refers to the ability to transform the tsunami of data into business.
Veracity
= quality.
Are the data accurate and clean?
Variety
Different (unstructured) data types.
The importance of different sources of information varies depending on the nature of the business.
Velocity
Continued stream.
Companies need the information flow quickly - as close to real-time as possible.
The data must be available at the right time to make appropriate business decisions.
Volume
Large amounts. This is at the base of Big Data.
5 V’s of Big Data
- Value
- Variety
- Veracity
- Velocity
- Volume
Why is data growing (exponentially)?
- Because data is generated in real-time
- Because data storage is becoming cheaper
- Because storage limits become bigger
- Because there is a move from analog storage to digital storage.
Unstructured data (meaning)
Information that is not arranged according to a pre-set data model or scheme.
What kind of data is unusable for analyses or estimations?
Unstructured data
Why is there more and more data available?
Due to digitalization; a lot of data is being generated in real-time.
Online customer sentiment
Online. Social-media based.
Strength customer satisfaction (2)
- Structured information.
- Control over who to ask.
Weaknesses customer satisfaction (2)
- Costly and time-consuming to collect.
- Many non-responses
Strengths online customer sentiment (2)
- Actual outspoken opinions
- Continuous (real-time) feedback
Weaknesses online customer sentiment (2)
- Not representative (very dissatisfied and satisfied people are more likely to say something)
- Less straightforward to analyze.
Microsoft Azure
Tool that analyses pictures, videos, and audio.
The tool can be used to provide information, for example, about the gender, age, hair color, facial hair, etc. of a person.
Similar tools by MA are also available that allow you to see if two different pictures contain the same person.
What is a problem occurring when offline media are becoming personalised?
To make offline ads personalised, information must be gathered, e.g., by scanning the faces, cars, or other of the person looking at the ad.
This can lead to PRIVACY CONCERNS.
Topic analysis
A machine learning technique that automatically assigns topics to text data.
Herewith, businesses are able to sift though large amounts of data quickly and pinpoint the most frequent topics mentioned in (e.g., customer feedback).
Machine learning method
Feeding the computer all kinds of pre-classified text. The algorithm then tries to determine certain patterns in the data. This can also be used to do a sentiment analysis.
Lexicon-based method
Also known as sentiment analysis.
Using a dictionary to classify words and text as positive or negative.
