User Tools

Site Tools



This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
projects:year3:14.2 [2019/08/21 14:22]
sally.johnson [Table]
projects:year3:14.2 [2019/08/22 10:22] (current)
sally.johnson [Project - Documents]
Line 4: Line 4:
 **Objectives:** **Objectives:**
-The objective of this project is to develop a framework for hotspot analysis. To accomplish thiswe (acreated a new approach to detect hotspots from point data, (b) developed a MapReduce approach to improve scalability of hotspot detection(c) created a new ensemble-based approach to hotspot prediction and (d) designed a MapReduce framework for ensemble-based prediction.+  * Utilize data from multiple source systems for analysis purposesby integrating both internal data and large volumes of data from outside systems (third party, social media). This integration empowers internal business users with the capability of analyzing customer info, product info, financial info, procurement info, weather data, etc. from multi-dimension and multi-view perspectives in different aggregation and granular levels. 
 +  * Develop effective and efficient data mining modelstools and techniques to improve business intelligenceby addressing real-world issues such as customer opinion, improving personalized service, and reducing customer attrition and improving cross-selling and up-selling in customer relationship management.
 </WRAP> </WRAP>
Line 16: Line 17:
 ===== Project - Impact and Uses/Benefits ===== ===== Project - Impact and Uses/Benefits =====
 <WRAP leftalign box > <WRAP leftalign box >
-Early hotspots detection and precise prediction are important in many disciplinesThis project provides scalable spatio-temporal hotspot detection and prediction approaches which are well engineered for big data. The approaches are general and can be applied in various application domains with minimal customizations. +We developed the proposed hybrid HDP-LDA model, and it can improve the performance of sentiment analysis in e-Business application. We tested our prototype system in many open data sets, however, the approach can be easily generalized to the IAB members' data setsThe outcomes of the project provide techniques for easily processing big data in analytic environments. The results of the study improve productivity for extracting greater value from big unstructured data. The modules of the system are implemented in Java languagesThe complete API will be provided in a software package.
- +
-This system can be beneficial in several applications such as epidemiology, public health, law enforcement, anti-terrorism, marketing (Merging markets, cross selling, and advertisement optimization), etc.+
 </WRAP> </WRAP>
Line 26: Line 25:
 ===== Project - Documents ===== ===== Project - Documents =====
projects/year3/14.2.1566415373.txt.gz · Last modified: 2019/08/21 14:22 by sally.johnson