TechnologyWhen we talk about analyzing vast amounts of consumer-generated media in order to find the “gold” in specific topics, issues, trends, opinions and sentiment, we’re talking about some powerful technologies that do the heavy lifting. Nielzen BuzzMetrics’ content mining capabilities are rooted in machine-learning and natural language processing technologies that mine unstructured data—vast amounts of raw text—to discover the intelligence it contains. These technologies are able to identify key phrases and words, detect the nature and strength of sentiment in text, classify and categorize data to provide meaning and relevance, and extract specific facts and data points to create the meaning and context that lead to intelligence. Nielsen BuzzMetrics technology can be trained to identify and analyze relevant messages about your brand, products and company from a variety of online external and internal sources. In addition to measuring the sheer volume of buzz, cpecific technologies include: Relevance DetectionThe ability to identify messages that are relevant to a specific goals/objectives of the project and to separate them from messages that are usually incidental references, etc. This feature can be used to rule out ambiguous or irrelevant messages that usually add "noise" to the project. ClassificationThe ability to identify pre-defined "topics" or "categories" of information in unstructured text. This feature can be used to classify messages into various topics that are relevant to the client's product, subject matter, company, issue, etc. and/or competitor(s). 
Phrase MiningThe ability to discover new, popular, different, or unique keywords and phrases from static data or a continuously streaming collection of unstructured data. This feature is very useful for detecting new "topics" or "categories" to add to the Classification scheme. 
Sentiment MiningThe ability to identify polar expressions (positive or negative) in unstructured data. The Sentiment Miner scans raw text for statements or subjective language that might indicate the author's opinions about a topic, brand, issue or company. Combined with Intelliseek’s topic detection technology, the Sentiment Miner can deliver a sense of "emotionality" and opinion on topics vital to clients' needs. 
Quote MiningThe automatic ability to generate positive and negative quotes from unstructured data, narrowing information by "topic" or "category," if needed. 
Concept MiningThe automatic ability to identify "concepts" in unstructured data, as well as their similarities, inter-relationships or connections. 
Social Network AnalysisThe automatic ability to identify users, interactions and their dynamic networks of influence in online communities. Which consumers are most vocal? Which are most influential? Which are most likely to be listened to by other consumers within the network? 
Fact ExtractionThe ability to extract relationships between various topics and to determine sentiment expressions in the text. This feature can be used to generate "facts" or "insights" that are relevant to the client's product, subject matter, company, issue, etc. and/or competitor(s). DispersionThe ability to identify where conversations take place and how they are dispersed, or how they spread, among other sources and segments Link AnalysisThe ability to identify Web sites, links and URLS that are mentioned and included in consumer-generated media |
BlogpulseTechnology at Work
CareersWe are looking for professionals seeking a challenging career in software development, marketing, sales, business development, management, or support. If you thrive on excellence and love challenges, explore opportunities with us today! Newsletter Sign Up |
 |
A subsidiary of The Nielsen Company