Dashboard showing sentiment results

Supporting Business Intelligence with A.I.

September Article 2019

Melbourne Business Analytics Conference 3rd of September 2019

Live sentiment analysis conducted on the day of the conference

Hello everyone. David from Data Oracle here.

Well, the conference was a great day out. There were 4 of us supporting the Data Oracle stand and our team did a great job managing the flood of interest in our service. It was a very successful day for us with many new business contacts established.

Anyway, we thought we’d conduct a live sentiment analysis at our stand on the day. We developed a simple survey as follows:

  • First name
  • Time of arrival
  • How would you describe the traffic during your commute this morning? Use positive, negative or indifferent words. Be descriptive.

We then tokenised each surveyed person’s words, removed stop words (a, I, and etc.), stemmed the words and ran the phrases through a sentiment algorithm. The sentiment algorithm evaluates the words in the phrases against a library of words which are assigned a positive or negative value. The algorithm then evaluates the overall sentiment value for the given phrase. 67 people were surveyed. Figure 1 below shows the data in the form of a Tableau dashboard. You can play with the live dashboard at Tableau Public by Clicking Here.

Figure 1

The chart of large dots shows the sentiment values of each participant against their time of arrival. The box and whisker plot shows the descriptive analytic data for all of the participants. The word map shows the top 30 words by highest frequency used by the participants.

Some key observations:

Sentiment Chart

  • The large majority of the participants arrived at the conference by 8:30am.

Box and Whisker Plot

  • The median value of all sentiment was zero.
  • 50% of the participants surveyed did not exceed sentiment ranges of between -0.026 to 0.125.
  • 3 people stood out as being very happy with their commute: Patsy, Daniella and Mellikaljlih.
  • 3 people stood out as being very unhappy with their commute: John, Shailee and Diana.

Word Map – top 4 observations

  1.     Traffic was the biggest issue for people; “traffic” being mentioned 26 times.
  2.     10 people mentioned the “train”, hence we assume they caught the train to the event.
  3.     8 people mentioned the word “smooth”.
  4.     7 people mentioned the word “positive”.

We thank everyone that participated and we hope you enjoyed the event and this short article.

Regards,

David

Director

Please follow and like us:

Leave your thought here

Social media & sharing icons powered by UltimatelySocial