Silent Day Detection on Microblog Data
TECHNOLOGY DESCRIPTION: This technology usesphrased-based weighted information gain (PWIG) and local query term coherence(LQC) to improve current silent day detectors. These methods weigh search termsdifferently either dependent on the order or the words (PWIG) or independent ofthe order of words (LQC), taking into account features such as the length ofthe search term and result popularity.STAGE OF DEVELOPMENT: Research and prototype testingare underway with reproducible results.APPLICATIONS: A process has been designed to isolate desired information onsocial media in order to meet the user’s interests. The user is able to findrelevant information posted over a long period of time and receive real timeupdates on it. PROBLEMS ADDRESSED: With an increase in the amount of information on social media,searching for relevant information can be difficult and overwhelming, sometimesgiving most recent and unrelated results. Results that aren’t relevant areknown as “silent day”, and detecting these is presently not ideal. Currentsearch technology focuses on short term searches, whereas this technology isable to evaluate long term searches. This technology better’s silent daydetection, thus assisting users in retrieving the information needed whiledisregarding irrelevant information.
United States
Patent Pending
美国
