Usamos cookies y otras tecnologias similares (Cookies) para mejorar su experiencia y proporcionarle contenido y anuncios relevantes para usted. Al utilizar nuestro sitio web, acepta el uso de Cookies. Puede cambiar su configuracion en cualquier momento. Politica de cookies.

Sentiment Analysis: More than a Good Idea Monitoring and Managing Perceptions in Social Media

Originally published septiembre 10, 2009

People have always discussed their favorite shows, stores, restaurants, products, etc. Until recently, marketing professionals and brand managers were not privy to those two-way conversations. Now, one-to-many conversations are taking place online at a furious rate. That opens the door to unprecedented opportunities for organizations to manage and strengthen brand reputation.

Print and broadcast outlets have always worked to produce fresh material on a predictable basis. People knew when new issues of magazines and newspapers would be published and when broadcast shows would appear on the air. Public feedback was usually limited to letters to the editor and calls to talk show hosts. We live in an “always on” world today. News outlets maintaining online sites that feature real-time updates and consumer generated material (e.g., blogs, videos, podcasts, personal reviews and Twitter) have become commonplace. Now, more than ever before, it’s the organizations that are effective at monitoring, participating and managing online activity that will succeed in understanding and responding to shifts in consumer preferences and demand.

The days of marketers meticulously analyzing each and every media mention are also a thing of the past. There’s just too much content. Marketing executives need the ability to monitor and react to a steady stream of real-time mentions from a host of disparate sources such as traditional media, the general public, customers, competitors, analysts and employees.

Today’s natural language processing technologies use the meaning of text to distill relevant information from massive amounts of online data, making sentiment analysis possible. Sentiment analysis software can give organizations a powerful tool for successfully dealing with today’s non-stop activity on Twitter, Facebook and other Internet sites. These solutions enable businesses to:

  • Track overall and feature-level sentiment toward a brand, product and organization.

  • Monitor and interact with customers regarding their experiences.

  • React quickly to negative items, to minimize damage.

  • Identify and address market trends.
Beyond monitoring what’s being said online, companies can also assess the quality of the information that is being posted. For example, a brand manager for a company may want to gauge the overall consumer attitude toward a new model. Analyzing product language or feature language provides a snapshot of what people are writing about that model. But it is often not accurate to say that an entire article or product review is either good or bad. One portion of the item may include positive comments while other references are negative. Or, if two products or brands are mentioned in the same post, it is important to know which brand is mentioned positively and which is referred to in negative tones.

More than labeling posted pieces as good, bad or neutral, sentiment analysis tools compute the entire document for both overall tone as well as granular evaluation for individual products and features. Sentiment analysis solutions make distinctions that ensure words and phrases are analyzed in the proper context. Words that are commonly categorized as a positive or negative can be used in the opposite tone. For example, the word “best” is normally a positive term. However, it takes on a different meaning in a sentence such as “you could do better.” These distinctions are critical.

Due to the volume and speed at which new content is being published in today’s information-centric world (and there seems to be no end in sight to the proliferation of social networking sites and user generated content), the monitoring task must be performed automatically.

Within hours, or possibly minutes, a company’s reputation can be influenced by the information and commentary being exchanged online. Issues need to be detected as fast as possible in order for an organization to react swiftly and appropriately. Sentiment analysis technology can process large volumes of documents very quickly. It is critical for marketing organizations to use real-time sentiment analysis to protect and strengthen their company’s brand equity.

SOURCE: Sentiment Analysis: More than a Good Idea

  • Manya Mayes
    Manya, a 14-year SAS veteran, plans and executes SAS' worldwide marketing strategies for text mining as SAS' Chief Text Mining Strategist. She has more than five years of text and data mining consulting experience with major manufacturers, financials, telco and retail, helping them better understand the information gathered through warranty, call centers, chat rooms and other customer data sources. She also has played an instrumental role in developing text mining applications for emerging technologies in the area of early warning, consumer and product safety in manufacturing, and most recently audio analysis. Manya is the author of a number of white papers and articles on text mining, and her work with SAS Text Miner began with its inception in 1998. Manya earned a Bachelor's degree in Mathematics from the University of Waikato, New Zealand, in 1991.

Related Stories



Want to post a comment? Login or become a member today!

Be the first to comment!