4 min read
Evaluating state-of-the-art sentiment analysis SaaS and open-source ML models on our movie review dataset.
Sentiment analysis (or opinion mining) is a natural language processing (NLP) technique used to determine whether data is positive, negative, or neutral. Sentiment analysis is often performed on textual data to help businesses monitor brand and product sentiment in customer feedback, and understand customer needs. Sentiment analysis requires creating and/or choosing different NLP/ML models, and testing and comparing different models/SaaS solutions. The task of finding and evaluating a solution can take not only weeks but months, but by using Tiyaro the effort is reduced to minutes.
Accelerating the rate of model evaluation using: Tiyaro Experiments
Tiyaro Experiments allows us a quicker way to compare different pre-trained machine learning models along with the SaaS Vendor APIs helping us accelerate the evaluation process. One such Sentiment Analysis of Customer Reviews for Movies Experiment we conduct is to detect sentiments of moviegoers using sentiment analysis models. For that we utilize the following State of the Art pre-trained machine learning models / SaaS API :
- twitter-xlm-roberta-base-sentiment
- Azure Cognitive Sentiment Analysis API
- sentiment-roberta-large-english
We start by searching for the sentiment analysis models /SaaS service, we can demo or directly consume the API. The model card also allows users to demo particular models. 
We can start the experiment by going to experiments and creating a new experiment. 
We need to select the type of experiment we are doing. Here we select "Text Classification."
We can add descriptions to our experiment, like metadata about our dataset, and the motivation behind the experiment.
We can select different models/SaaS services to evaluate our dataset on, we can add those from our bookmarks models. We need to add datasets to run and evaluate our experiment.
We can now 🏃♀️ our experiment!
Conclusion
We will get an email notification after the experiment has finished now we can analyze the experiment.
We can download the results as a JSON object and perform further analysis as required. We also get different confidence scores along with the inferencing result for our experiment. We can quickly evaluate which model is better to use for our particular use case.
You can share the experiment with your coworkers and on social media. Also, you can make a copy of the given experiment to enhance or modify the particular experiment.
Wish to create one? Head on over to Tiyaro !