Search
Program Calendar
Browse By Day
Browse By Time
Browse By Person
Browse By Room
Browse By Division
Browse By Session Type
Search Tips
Personal Schedule
Sign In
In this paper, we introduce a novel method for analyzing entity framing in a large corpus of online news articles. This measure is proposed as a solution to the heretofore intractable problem of large-scale analysis of framing for particular entities within corpora too large to be reliably hand-coded and for which no document selection procedures have been devised. In this method, a semantic network is constructed from preprocessed text wherein nodes represent individual words/n-grams, and edge weights represent cosine distance between words or n-grams in high-dimensional vector space. A random walk with restart (RWR) is initiated across the graph from the node of interest, creating a list of words associated with the node which can then be analyzed using a variety of computational or manual coding techniques. We demonstrate that this method serves as a helpful tool for the analysis of latent frames for both media researchers and journalists.
Jacob T Fisher, The U of California, Santa Barbara
Devin J. Cornell, U of California, Santa Barbara
Frederic Rene Hopp, U of California, Santa Barbara
Rene Weber, U of California - Santa Barbara