Video Sentiment Analysis

1
(No Transcript)
2
How to do YouTube Sentiment Analysis for
brand-insights

• www.repustate.com

3
Sentiment Analysis For Youtube Videos

• YouTube sentiment analysis can be very valuable
  for brand insights. In this blog, we discuss how
  you can search, find, and retrieve insights from
  hundreds of YouTube videos with Repustates video
  analysis tool. We also broadly explain how
  sentiment analysis for YouTube comments is done.

• Marketing strategy in the age of social media
  listening includes uncovering brand and customer
  insights from YouTube videos. There are virtually
  millions of feelings and opinions about brands on
  YouTube everyday, expressed by people across all
  ages.

4
What is aspect-based sentiment analysis of video
reviews?

• Aspect-based sentiment analysis breaks down a
  review into smaller segments, and studies them
  for sentiment, thus enabling more detailed and
  accurate insights. Aspect-based sentiment
  analysis can easily help distinguish which
  features of a product or service are liked and
  which ones can be improved.
• Lets see an example
• Went to Bar Chef last night and loved their
  drinks, especially the martinis, but the food was
  horrible. My nachos tasted microwaved and the
  calamari was rubbery.
• This review needs to be analyzed at the aspect
  sentiment level, with further aspect insights on
  Drinks (martinis), and Food are revealed through
  the aspects of nachos and calamari.

02
03
5
How does Repustates video analysis tool perform
YouTube sentiment analysis?
02
03
6
Repustates video content analysis tool conducts
aspect based sentiment analysis on YouTube videos
to deliver the most granular brand insights. It
uses advanced named entity recognition (NER) to
identify named entities in YouTube videos and
classifies them into predetermined categories.
NER classifies company names, geo-locations,
things, and names of people who are mentioned in
the videos. These insights thus can be used to
improve marketing efforts, products, customer
experience, or customer service.
02
03
7

• Step 1 Collect prepare video/audio/image/text
  dataVideos are converted into text using
  speech-to-text transcription models and run
  through neural networks (NN) for audio content
  analysis. These NNs also discover caption
  overlays in videos, and if detected, they read
  and extract text from it. They also employ image
  detection for logos in background imagery.All
  this video data, along with text data from the
  comments is collected and manually edited to
  remove redundancies, punctuations, gifs, emojis,
  etc. It is then converted in a machine-readable
  format (CSV, XLS, JSON) so it can be ingested
  into the machine learning pipeline for training.

02
03
8

• Step 2 Apply sentiment analysisThe data is run
  through the sentiment analysis API for opinion
  mining. It quickly returns sentiment scores for
  each relevant topic, aspect, or entity ranging
  from -1 for negative emotions, 0 for neutral
  feelings, and 1 for positive sentiment.
• Step 3 Visualize insightsSentiment scores are
  presented in the form of visual reports
  consisting of charts, graphs and tables through a
  sentiment visualization dashboard.

02
03
9
How is Sentiment analysis for YouTube comments
done?

• YouTube comments analysis can help with vital
  insights for media monitoring not just for
  products and services but also for corporate and
  individuals in key positions. Sentiment analysis
  for YouTube comments is done in broadly 3 steps
• Step 1 - Scrapping Preparing Youtube comments
• Step 2 - Running it through Sentiment analysis
  API
• Step 3 - Data visualization

02
03
10
Thank you!

• Understand your data, customers, employees with
  12X the speed and accuracy.

• Visit www.repustate.com to learn more