Communications of the ACM
Google's Language Techniques Help O2 Czech Republic Reveal Network Secrets
The Word2vec neural network technique developed to understand human languages, can also interpret raw cell tower data.
Credit: lifestyletrading101.com
Word2vec, a neural network technique developed by researchers at Google to understand human languages, can also interpret raw cell tower data, a breakthrough that could potentially improve network performance, according to researchers at O2 Czech Republic.
The independent network provider hopes to develop the technique to find trends in customer geolocation, and to overcome the problem of unreliable data resulting from SIM cards connecting to network base transceiver stations.
Word2vec is a group of machine learning models that express words as vectors, normally in 100 or more dimensions, based on analysis of a corpus of data, such as the text from Wikipedia.
The O2 Czech Republic researchers used Word2vec for each cell, creating a 100-dimensional vector for each of 50,000 cell IDs. Then, they reduced the number of dimensions to produce a meaningful interpretation of the data.
From ZDNet
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA
No entries found