| Presentation | 2008-06-30 Preference Segmentation Based on User Search History Yusuke Fukazawa, Takefumi Naganuma, Shoji Kurakake, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Behavioral targeting method, which delivers advertisement/information according to preference segmentation estimated by user's web activity, has been paid attention intensively. The construction of estimation-rule of preference segmentation is the key to realize high precision behavioral targeting. In a past, estimation-rule has been built based on subject interview; however, there are a couple of possible problems. One of them is to collect the estimation-rule from non-target subject because of the difference of the way subject and architect think the area of the preference each segmentation holds. To solve this problem, we utilize generalized task library, which we constructed with Mizoguchi-Lab at Osaka univ. in order to define the action broadly for every preference segmentation. The other problem is that the ambiguousness of the criteria to collect web activity history for each preference segmentation causes the collection of which cannot identify one preference segmentation. We solve this problem by excluding the web activity history collected with non-defined action. Finally, we evaluate the efficiency of our method from the viewpoint of coverage of the definition of the action and validity of the web activity history collected. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Behavioral targeting / Preference estimation / Information retrieval |
| Paper # | AI2008-7 |
| Date of Issue |
| Conference Information | |
| Committee | AI |
|---|---|
| Conference Date | 2008/6/23(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 | Artificial Intelligence and Knowledge-Based Processing (AI) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Preference Segmentation Based on User Search History |
| Sub Title (in English) | |
| Keyword(1) | Behavioral targeting |
| Keyword(2) | Preference estimation |
| Keyword(3) | Information retrieval |
| 1st Author's Name | Yusuke Fukazawa |
| 1st Author's Affiliation | NTTDoCoMo() |
| 2nd Author's Name | Takefumi Naganuma |
| 2nd Author's Affiliation | NTTDoCoMo |
| 3rd Author's Name | Shoji Kurakake |
| 3rd Author's Affiliation | NTTDoCoMo |
| Date | 2008-06-30 |
| Paper # | AI2008-7 |
| Volume (vol) | vol.108 |
| Number (no) | 119 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |