MeaningCloud provides entirely customizable sentiment analysis to get maximum accuracy
Its new customization tools enable users to carry out a sentiment analysis whose quality is comparable to what human analysts would achieve
19 APRIL 2016, NEW YORK
SummaryMeaningCloud, a leading text analytics products provider, has announced that has incorporated into its sentiment analysis tools new customization capabilities that enable users to perform an opinion analysis fully adapted to their domain. These capacities are superior to those currently available on the market, since they allow, besides associating words with polarities, defining these polarities taking into account modifiers, the context, and the function of every word at every moment.
MeaningCloud, a leading text analytics products provider, has announced that has incorporated into its sentiment analysis tools new customization capabilities that enable users to perform an opinion analysis fully adapted to their domain. These capacities are superior to those currently available on the market, since they allow, besides associating words with polarities, defining these polarities taking into account modifiers, the context, and the function of every word at every moment.
Sentiment analysis is one of the key text analytics features because it permits to extract the polarity of the free-format opinions expressed in social media conversations, satisfaction surveys or contact center interactions. However, this application has always had to face a tough challenge: the ambiguity of language and its differences depending on the diverse contexts, which is reflected in the fact that an expression may have different meanings depending on the situation or linguistic register. For example, the word "cheap" generally has a positive connotation, unless we are talking about a luxury product, in which case it can be somewhat derogatory. The variants of a language spoken in different regions also intensify this problem.
Generic sentiment models for a language tend to be unable to collect all its possible variants and therefore sometimes do not provide enough quality in the analysis. For this reason, the adoption of these tools is relatively small, compared to their potential benefits.
A breakthough in semantic analysis
The new functionality for creating personal sentiment models presented by MeaningCloud enables companies to fully adapt the sentiment analysis to their domain, application or industry. These customization capabilities are based on MeaningCloud's powerful Natural Language Processing technology and allow users to develop autonomously - without programming - powerful sentiment analysis engines tailored to their needs.
Unlike other technologies available on the market, which essentially define "bags of words" with either positive or negative polarity, MeaningCloud's tools go far beyond and enable to
- Define the role of a word as a polarity vector (container, negator, modifier), allowing to use lemmas to easily incorporate all the possible variants of each word
- Specify particular cases of a word's polarity, depending on the context in which it appears or its morphosyntactic function in each case
- Define multiword expressions as priority elements in the evaluation of polarity
- Manage how these custom polarity models complement or replace the general dictionaries of every language.
For example, the expression "the interest rate is very high" expressed by a financial services customer may be positive if it refers to deposits, but negative if it has to do with mortgages. With this tool, it is possible to define these different polarities for each case.
The result of the contextualization of sentiment analysis is an excellent accuracy, very close to what would be achieved by human resources.
Custom sentiment models complement other customization capabilities of MeaningCloud, which include text classification models and topic dictionaries, which together provide the highest-quality text analytics to all developers.
See more here.
If you want to know more about MeaningCloud's sentiment analysis customization capabilities, join us in this webinar on May 4th.
"The adaptation to the domain is what makes the difference between a good sentiment analysis and an exceptional one. With these customization tools, we are making available to all developers sentiment analysis engines completely adapted to their context, which provide the highest-quality analysis." José C. González, CEO of MeaningCloud