It’s crucial for businesses to listen, measure, and understand what people are saying about them. Restaurants are no different, every day millions of reviews get posted on social media by the consumers. Sentiment analysis helps brands go beyond the vanity metrics such as Likes, Shares and Mentions to unveil the actual feeling of the consumer. Was it positive? Negative? Sarcastic?
Capturing the human emotion for a product or service helps businesses understand if the brand message resonates with the audience, and at the same time it also provides solid indications with respect to changes in the overall business strategy. Sentiment Search operates in this space and caters to the Restaurant industry by providing social media analytics platform.
It has been close to a year since PromptCloud started working with Sentiment Search to help them with their web data acquisition requirements for social media analytics. We recently caught up with Prithvi Dhanda, Founder of Sentiment Search, to gain a better understanding of how their solution works and why data scraping is essential to them.
Q1: Please tell us about Sentiment Search and explain the offerings to our readers?
Sentiment Search is built on two core technologies; Natural Language Processing, and Sentiment Analysis. We have developed two products using these technologies;
- Social Media Analytics: This platform is used to aggregate reviews for thousands of restaurants across multiple cities and provide comprehensive insights and search capabilities.
- Research Bot: This platform is developed for universities that can be used to search multiple research portals in order to identify relevant PDF files and highlight relevant text within these files.
Q2: How does the “Social Media Analytics” solution for restaurants work? What type of algorithms do you use for analysis?
The Social Media Analytics platform aggregates thousands of reviews and tweets from multiple sources for multiple cities on a daily basis. We use a range of sentiment analysis algorithms including, Naïve Bayesian algorithms, Neural Networks, and Rule-Based algorithms.
Q3: What are your biggest challenges when it comes to acquiring and analysing web data?
Our company requires specific content from specific web pages, therefore developing crawlers for each site is a tedious task, especially if the target site does not offer an API.
Q4: If you could go back in time, what advice would you give to your former self who is about to start the business?
Q5: What prompted you to choose PromptCloud? How did PromptCloud solve your business problem with the data extraction services?
I came across PromptCloud’s webscraping service while searching for data crawling companies on Google. PromptCloud offers a comprehensive API which has made the collection and processing of data relatively easy. Their customer support is also very reassuring.