| Presentation | 2011-03-28 Event Clustering Method based on Linear HMM for Geotagged Personal Contents Iku OHAMA, Takuya KIDA, Hiroki ARIMURA, Toshihisa ABE, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | We discuss the geographical clustering problem on human activity logs. In this paper, we formulate the problem as a segmentation problem for 2-dimensional sequence data, and propose two clustering methods, named LS-linHMM and X-linHMM. LS-linHMM is based on linear constrained HMMs which estimate the optimal number of clusters by model selection with an information criterion. X-linHMM is a hierarchical clustering method which is based on the idea of x-means and use 2-state linear constrained HMMs. It runs faster than LS-linHMM in many cases. In this time, we apply these clustering methods into the clustering problem for geo-tagged personal picture contents. Our experimental results using real data showed that the proposed methods are more effective in clustering than the simple x-means. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | personal contents / geo-tag / Hidden Markov Model / x-means / clustering |
| Paper # | IBISML2010-105 |
| Date of Issue |
| Conference Information | |
| Committee | IBISML |
|---|---|
| Conference Date | 2011/3/21(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Information-Based Induction Sciences and Machine Learning (IBISML) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Event Clustering Method based on Linear HMM for Geotagged Personal Contents |
| Sub Title (in English) | |
| Keyword(1) | personal contents |
| Keyword(2) | geo-tag |
| Keyword(3) | Hidden Markov Model |
| Keyword(4) | x-means |
| Keyword(5) | clustering |
| 1st Author's Name | Iku OHAMA |
| 1st Author's Affiliation | Corporate R&D Division, Panasonic Corporation() |
| 2nd Author's Name | Takuya KIDA |
| 2nd Author's Affiliation | Division of Computer Science, Hokkaido University |
| 3rd Author's Name | Hiroki ARIMURA |
| 3rd Author's Affiliation | Division of Computer Science, Hokkaido University |
| 4th Author's Name | Toshihisa ABE |
| 4th Author's Affiliation | Panasonic Advanced Technology Development Co., Ltd. |
| Date | 2011-03-28 |
| Paper # | IBISML2010-105 |
| Volume (vol) | vol.110 |
| Number (no) | 476 |
| Page | pp.pp.- |
| #Pages | 8 |
| Date of Issue | |