ResourcesResources

Demo

Documentation

Smart Sentiment icon 100x100

Use CasesUse Cases

  • News
    Find the current sentiment and add a positive spin to new Blog Posts.
  • Social Media Sentiment
    Track how your company is perceived and take action.
  • Product Reviews
    Correct or expand the reaction to your product.

Automated Sentiment Analysis

Heart Robot

We have an extensive compiled entity dictionary as part of our knowledge base, and we update it daily. Our knowledge base automates the NLP (Natural Language Processing) sentiment analysis for overall text, entities, and keywords.  By selecting a domain area such as news, product reviews, or social media, the service tailors to the domain and refines the sentiment accuracy. We have been refining our algorithms to reflect a wide spectrum of polarities. Natural Language Understanding (NLU) allows us to continually train our service to recognize the similarities within the text in the domain and become more accurate with the results.

Customized Dictionaries

Several customers have special entities and keywords particular to their business. We have modified our automated text analysis to allow for user-defined customer dictionaries for additional entities. When selected, the custom entity dictionary is used to identify and extract entities that are not in the current dictionary. You can update the dictionary as these new entities become part of your business. 

Bookshelf Ninja

Article Sentiment

 

Sentiment Analysis Demo Image

Smart Sentiment services assigns an overall sentiment: positive, negative, neutral, or no sentiment to the original text.  The sentiment is based on a weighted comparison of all sentiment words in the initial text document.  Sentiments are based on a weighted comparison of all natural language words denoting or evoking a sentiment, such as “good”, “bad”, “lovely”, “splendid”.

The “neutral” and “no sentiment” are not the same. “No sentiment” means there is not enough sentiment words to determine a sentiment.  A “neutral” sentiment results from the presence of both “positive” and “negative” sentiments that neutralize each other. 

Entity Sentiment

Smart Sentiment services extracts entities (persons, places, or organizations) from the text, and assigns a sentiment per entity. The sentiment uses the local context or sentiment words in the vicinity of the entity to help determine the sentiment score. We identify at least one category for every entity and multiple categories if they exist. The additional categories are possibilities for new audiences or business opportunities. 

Smart Sentiment - Entity

Keyword Sentiment

The keyword score shows the importance of the keyword to the original text. We assign a sentiment to each keyword or keyword phrase listed in the text. The sentiment score reflects the only the sentiment of this keyword in this text. The same keyword could have a different sentiment in another text because the sentiment is determined by the context. Having an individual keyword sentiment score, and an entity sentiment score, and overall score gives a more complete picture of the sentiment around your topics and categories.

Smart Sentiment Keyword Sentiment Analysis

Sentiment Segmentation

Smart Sentiment - Segmentation

The “Segmentations” tab shows how we divided the data with our automated services.  You can continue to refine the results to build out your application.

Have Questions?

Contact Us

Smart Sentiment Pricing

  • 2,000 requests per day
  • $0.0185 extra charge for additional requests over the daily allowed maximum
  • 5,000 requests per day
  • $0.0150 extra charge for additional requests over the daily allowed maximum
Read More
$250/month
Looking for more?
  • Contact us for a quote
  • No daily limit on requests
  • Easy to scale
  • Pay only for what you use
$0.019/Per Request