2016 Fiscal Year Annual Research Report
N-dimensional Data Mining based Evolutionary Customer Behavior Modeling
| Project/Area Number |
25330270
|
|---|
| Research Institution | Hosei University |
|---|
Principal Investigator |
Huang Runhe 法政大学, 情報科学部, 教授 (00254102)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
Jianhua Ma 法政大学, 情報科学部, 教授 (70295426)
|
|---|
| Project Period (FY) |
2013-04-01 – 2017-03-31
|
| Keywords | Knowledge discovery / data mining engine / Knowledge fusion engine / Knowledge depository / personalized model / smart service / decision-making |
|---|
| Outline of Annual Research Achievements |
In year 2016, our research was focused on
(1) Verifying and improving our business wisdom (BW) framework with n-dimensional big data. We used 2-year retail business data which is about 800000 data records with about 80 attributes; (2) Applying our KID model together with knowledge graph (in Neo4j framework) for knowledge repository and a inference mechanism (in JESS platform) for business decision-making and prediction; (3) Building individual customer personalized model with knowledge discovery about customers based on Cyber-I platform so as to serve each customer better (such as making prediction of their shopping behaviors, and pushing right advertisement and giving recommendation to them); (4) Extending our successful platform with proposed KID model (a cognitive approach to knowledge discovery) to other application fields such as city traffic generated big data for gaining knowledge about stream of people and traffic for smart system and robot's cognitive development and behavior; (5) Putting our effort to make the KID model as a generic cognitive process model towards the core element for machine intelligence.
To conclude, we have dedicated to our proposed research, sufficiently exchanged our ideas with the external researchers, and actively involved international academic society activities. In year 2016, we published 5 international journal articles, 4 international conference papers, 1 domestic conference paper, and 1 international edited book.
|
