| Presentation | 2017-03-03 Classification of Highly-Accurate Identifiable Applications Using Gini Index and Cosine Similarity Takamitsu Iwai, Akihiro Nakao, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Existing research on application identification has problems from three points of view; credibility of training data, flexibility of learning, and violation of privacy. Therefore, we propose a system that classifies mobile traffic using modified smartphones that send packets with application tags. This system has solved the problems mentioned above. We evaluate this system using a trace of real traffic and show this system can classify the 80% of the mobile only using the statistics of packets (e.g., the length of packets). We focus on applications that can be classified accurately using only destination IPs because they connect limited server. We propose the method that distinguishes these applications using Gini index and cosine similarity and classify mobile traffic sent by them accurately. We evaluate this method in real mobile traffic and show that we can classify 92% of about 14 applications traffic when learning period is set to 1 day. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | application identificaiton / machine learning / MVNO |
| Paper # | NS2016-192 |
| Date of Issue | 2017-02-23 (NS) |
| Conference Information | |
| Committee | NS / IN |
|---|---|
| Conference Date | 2017/3/2(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | OKINAWA ZANPAMISAKI ROYAL HOTEL |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | General |
| Chair | Hideki Tode(Osaka Pref. Univ.) / Katsunori Yamaoka(Tokyo Inst. of Tech.) |
| Vice Chair | Yoshikatsu Okazaki(NTT) / Takuji Kishida(NTT) |
| Secretary | Yoshikatsu Okazaki(Kyushu Inst. of Tech.) / Takuji Kishida(NTT) |
| Assistant | Shohei Kamamura(NTT) / Kunitake Kaneko(Keio Univ.) / Takashi Natsume(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Network Systems / Technical Committee on Information Networks |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Classification of Highly-Accurate Identifiable Applications Using Gini Index and Cosine Similarity |
| Sub Title (in English) | |
| Keyword(1) | application identificaiton |
| Keyword(2) | machine learning |
| Keyword(3) | MVNO |
| 1st Author's Name | Takamitsu Iwai |
| 1st Author's Affiliation | University of Tokyo(UTokyo) |
| 2nd Author's Name | Akihiro Nakao |
| 2nd Author's Affiliation | University of Tokyo(UTokyo) |
| Date | 2017-03-03 |
| Paper # | NS2016-192 |
| Volume (vol) | vol.116 |
| Number (no) | NS-484 |
| Page | pp.pp.199-204(NS), |
| #Pages | 6 |
| Date of Issue | 2017-02-23 (NS) |