Human dynamics through big data

Илтгэсэн хурлын нэр: Хүрэл тогоот- 2014

Илтгэгч:  Д.Золзаяа

Хамтран илтгэгч: [Д.Золзаяа:D.SW43],Stanislav Sobolevsky

Илтгэсэн огноо: 2014-10-25

Илтгэлийн хураангуй: Technology development leads billions of data that has been generated everywhere and every time. This enormous amount of data often called b i g data most important thing for delivering new insights to decision makers. It contains behavioral information and different types of human activity. Early recognition and prediction of human behaviors are of great importance in many societal applications like health-care, capacity planning, risk management and urban planning, etc. When this type of data coupled with relevant datasources like open geographical map available online (e.g., Openstreetmap, Google map, etc), the identification of human activities (e.g., working, studying, shopping, etc) by geographical areas can be enabled from mobile phone data. In this paper, we propose a methodology that profiles geographical areas by human activities for classifying mobile phone data into call frequency patterns. The call frequency patterns are semantic (high level) descriptions to call frequency in mobile network event. For profiling areas, we use unsupervised spectral clustering followed by k-means clustering algorithm based on TF/IDF cosine similarity metric. We demonstrate the methodology to interpret standard or exceptional(divergent) type of call frequency patterns and evaluate/validate our approach using a real dataset containing 1 million Call Detailed Records. This type of analysis and its application is important to understand hidden patterns and unknown correlations and other useful information that can be used to make better decisions.

Илтгэлийн төрөл: Дотоодод зохион байгуулагдсан олон улсын ЭШХ

Түлхүүр үг:

Илтгэл нэмсэн: Д.Золзаяа

Монгол Улсын Шинжлэх Ухаан Технологийн Их Сургууль © 2020