Trendkite Says: Media Monitoring? For Best Results Use NLP

  1 m, 15 s

TrendKite PicAs a PR analytics solution and Lexalytics partner, Trendkite knows how crucial accurate media monitoring is for your brand.

That’s why Trendkite founder Matt Alison explains that the key ingredient to more accurate media monitoring: Natural Language Processing.

In a post on Trendkite’s blog, Matt breaks down how most media monitoring solutions work. They use Boolean search to sift through millions of documents and return relevant results. Boolean search combines keywords with operators like AND, NOT, and OR to narrow down results and return only the most relevant results.

“For example, a Boolean search could be “taxi service” AND “San Francisco”. This would limit the search results to only those documents containing the two keywords.”

But when it comes to media monitoring, there are some challenges that Boolean search alone can’t tackle – synonyms and polysemy. When many words refer to the same thing, like with synonyms, or one word has more than one meaning, the definition of polysemy, Boolean logic isn’t enough.

The solution? Natural Language Processing. NLP, based on machine learning, is the ability of software to understand and parse speech like a human would. Text mining software like Salience uses NLP to analyze text and return useful and relevant results.

Matt concludes, “The best PR solutions add sophisticated big data analysis to natural language search to give users clarity and insight.”

We couldn’t agree more.

For more, read Matt’s original post on the Trendkite blog – How Accurate is Your Media Monitoring

Categories: Natural Language Processing, Partners