Friend recommendation (FR) becomes a valuable service in location-based social networks. Its essential purpose is to meet social demand and demand on obtaining information. The most of current existing friend recommendation methods mainly focus on the preference similarity and common friends between users for improving the recommendation quality. The similar users are likely to have similar preferences of point-of-interests (POIs), the kinds of information they provided are limited and redundant, can not cover all of the target user’s preferences of POIs. This paper aims to improve amount of information on users’ preferences through FR. We give a definition of friend recommendation considering preference coverage problem (FRPCP), and it is also one NP-hard problem. This paper proposes the greedy algorithm to solve the problem. Compared to the existing typical recommendation approaches, the large-scale LBSN datasets validate recommendation quality and significant increase in the degree to preferences coverage.
|Original language||English (US)|
|Title of host publication||Advances in Knowledge Discovery and Data Mining - 21st Pacific-Asia Conference, PAKDD 2017, Proceedings|
|Editors||Longbing Cao, Kyuseok Shim, Jae-Gil Lee, Jinho Kim, Yang-Sae Moon, Xuemin Lin|
|Number of pages||15|
|State||Published - 2017|
|Event||21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017 - Jeju, Korea, Republic of|
Duration: May 23 2017 → May 26 2017
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Other||21st Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2017|
|Country/Territory||Korea, Republic of|
|Period||5/23/17 → 5/26/17|
Bibliographical noteFunding Information:
This work was supported in part by the National Science Foundation grants IIS-61370214, IIS-61300210.
© 2017, Springer International Publishing AG.
- Friend recommendation
- Power-law distribution
- Preference coverage