Discovery of significant emerging trends

Saurabh Goorha, Lyle H. Ungar

Abstract

We describe a system that monitors social and mainstream media to determine shifts in what people are thinking about a product or company. We process over 100,000 news articles, blog posts, review sites, and tweets a day for mentions of items (e.g., products) of interest, extract phrases that are mentioned near them, and determine which of the phrases are of greatest possible interest to, for example, brand managers. Case studies show a good ability to rapidly pinpoint emerging subjects buried deep in large volumes of data and then highlight those that are rising or falling in significance as they relate to the firms interests. The tool and algorithm improves the signal-to-noise ratio and pinpoints precisely the opportunities and risks that matter most to communications professionals and their organizations.
BibTeX
@inproceedings{goorha2010discovery,
  author = {Saurabh Goorha and Lyle H. Ungar},
  title = {{Discovery of significant emerging trends}},
  booktitle = {ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
  year = {2010},
  doi = {10.1145/1835804.1835815},
}

Loading related content...

Figure