SENTI

Inspiration

We decided to create this application because we consider it imperative for a company to get to know what people say about them and their products. By using this application, the company can take advantage of the data to adapt and optimize their business in a way that suits their customer base.

What it does

Takes data from specific sources and analyzes them according to the sentiment. An additional function of our SENTI app is users are able to compare data obtained from searching different keywords e.g. a business competitor. They also have the possibility to see exact statements where both of the searched keywords are mentioned.

How we built it

Process We tried to find out what people think about a company by using keywords as search terms. We used social media platforms as the data source.

1. Our data source was Twitter, where we extracted tweets where the keyword “Coca Cola” was mentioned.

2. We used the Logic App to extract data from twitter. We also used Cognitive Services, more specifically text analysis to extract the sentiment of the tweets and key phrases. Furthermore we employed Translator to translate tweets where the original language was not English.

3. The raw data that was extracted is stored in the Cosmos Database in Microsoft Azure.

4. We used AI to process the data. We used the Text Analytics service which is a cloud-based service that provides advanced natural language processing over raw text for sentiment analysis, key phrase extraction, named entity recognition, and language detection. Sentiments are used to find the ratings of the reviews by assigning a score from 0 to 1. Zero represents an overall negative sentiment while one represents a positive sentiment.

5. This processed data is stored in a Cosmos cloud database.

6. We connected the Cosmos database to Power BI in order to create diagrams and graphs that help us visualize the data.

Challenges we ran into

We did not have any experience working with Azure or Power Bi so we had to learn everything from scratch and encountered a number of issues along the way.

As with most projects it is always hard to agree on a specific design for a product. It takes some time to get all members to agree on one design.

Accomplishments that we're proud of

We are proud of the prototype that we created. It has the functionality that we planned and given how inexperienced we were when we started, this makes us very happy. We are also proud of the design of the product as it’s simple and intuitive.

What we learned

With this Hackathon we learned how to use AI in practice, improved our skills with databases and frontend skills. We also worked with Power BI for the first time. We became familiar with Azure.

What's next for SENTI

Hopefully real implementation and usage of the product. There are still a lot of possibilities on how to extend this application.

Built With

• azure
• cognitiveservices
• cosmosdatabase
• figma
• logicapp
• office365
• powerbi