| Description | Web users worldwide rely on search engines daily, querying diverse terms to locate pertinent information. Due to the omnipresence of search engine in contemporary lives, we hypothesise that finely grained analyses of these search terms volume can offer valuable insights into societal trends, potentially reflecting economic conditions and overall quality of life.
The present project will examine search engine trends, sampled daily and hourly, for dozens of specific search terms and attempt to reveal temporal and spatial patterns in how these keywords are used across daily search activities.
In a preparatory stage, the student(s) will develop / refine datamining protocols and make use of machine learning methods such Digital Time Warping and Hierarchical Clustering to categorise time-series.
Once this is achieved, Student(s) will implement Deep Learning methods to test how effective learning of the categorisation is achieved. The original solutions used at this stage could range from the use of Long Short Term Memory RNN, Time Series Transformer, Attention-based RNN or even to Hidden Markov models (Each of these alternatives has its own strengths depending on the specific characteristics of your time series data and the nature of the categorization task and there will be trade offs to examine) |
|---|