Journal of the Korea Convergence Society (한국융합학회논문지)

Korea Convergence Society (한국융합학회)
권호 표지
DOI QR코드

DOI QR Code

Natural Language Processing-based Personalized Twitter Recommendation System

자연어 처리 기반 맞춤형 트윗 추천 시스템

  • 이현창 (광운대학교 컴퓨터과학과) ;
  • 유동필 (광운대학교 컴퓨터과학과) ;
  • 정가빈 (광운대학교 컴퓨터과학과) ;
  • 남용욱 (광운대학교 컴퓨터과학과) ;
  • 김용혁 (광운대학교 컴퓨터과학과)
  • Received : 2018.09.13
  • Accepted : 2018.12.20
  • Published : 2018.12.28
Download PDF
⟨ Previous Next ⟩

Abstract

Twitter users use 'Following', 'Retweet' and so on to find tweets that they are interested in. However, it is difficult for users to find tweets that are of interest to them on Twitter, which has more than 300 million users. In this paper, we developed a customized tweet recommendation system to resolve it. First, we gather current trends to collect tweets that are worth recommending to users and popular tweets that talk about trends. Later, to analyze users and recommend customized tweets, the users' tweets and the collected tweets are categorized. Finally, using Web service, we recommend tweets that match with user categorization and users whose interests match. Consequentially, we recommended 67.2% of proper tweet.

트위터 사용자는 팔로우, 리트윗 등을 사용하여 자신이 관심 있어 하는 트윗을 찾는다. 하지만 사용자가 3억여 명에 달하는 트위터에서 사용자가 관심 있는 트윗을 찾기는 힘든 일이다. 이를 해결하기 위해 본 논문에서는 사용자 맞춤형 트윗 추천 시스템을 개발하였다. 우선, 사용자에게 추천할 수 있을 만한 가치가 있는 트윗을 수집하기 위해 현재 트랜드를 수집하고, 트랜드에 대해 이야기하는 인기 있는 트윗들을 수집한다. 이후 사용자를 분석하고 맞춤형 트윗을 추천하기 위해 사용자의 트윗과 수집한 트윗을 범주화한다. 최종적으로 웹서비스를 이용하여 사용자에게 본인과 카테고리가 일치하는 트윗과 관심사가 일치하는 사용자를 추천해준다. 결과적으로 67.2%로 적절한 트윗을 추천하였다.

Keywords

OHHGBW_2018_v9n12_39_f0001.png 이미지

Fig. 1. Example of word2vec in the vector space. (a) Vector offsets for three word and (b) Example of word placement in vector space

OHHGBW_2018_v9n12_39_f0002.png 이미지

Fig. 2. System arcitecture of our tweet recommendation

OHHGBW_2018_v9n12_39_f0003.png 이미지

Fig. 3. Pre-processing of “나무위키”. (a) Original data for “나무위키” and (b) Results of preprocessing

OHHGBW_2018_v9n12_39_f0004.png 이미지

Fig. 4. Example of konlpy

OHHGBW_2018_v9n12_39_f0005.png 이미지

Fig. 5. Flow chart of our keyword extraction

OHHGBW_2018_v9n12_39_f0006.png 이미지

Fig. 6. Example of morphology tweet analysis

OHHGBW_2018_v9n12_39_f0007.png 이미지

Fig. 7. Example of calculating WM and TDM

OHHGBW_2018_v9n12_39_f0008.png 이미지

Fig. 8. Example of our Web service

OHHGBW_2018_v9n12_39_f0009.png 이미지

Fig. 9. Example of tweet analysis function in our Web service

OHHGBW_2018_v9n12_39_f0010.png 이미지

Fig. 10. Example of personal information management

OHHGBW_2018_v9n12_39_f0011.png 이미지

Fig. 11. Example of tweet Recommend function in our Web service

OHHGBW_2018_v9n12_39_f0012.png 이미지

Fig. 12. User matching based on user interest

Table 1. Example of similarity comparison of tweets and keywords

OHHGBW_2018_v9n12_39_t0001.png 이미지

Table 2. Example of our keyword extraction

OHHGBW_2018_v9n12_39_t0002.png 이미지

Table 3. List of information that can be obtained after Twitter login

OHHGBW_2018_v9n12_39_t0003.png 이미지

References

  1. D. M. Boyd. & N. B. Ellison. (2007). Social network sites: Definition, history, and scholarship. Journal of computer-mediated Communication, 13.1, 210-230. https://doi.org/10.1111/j.1083-6101.2007.00393.x
  2. J. S. Min. (2012). Study on Twitter users' political participatio. Journal of Communication Science, 12.2, 274-303.
  3. S. H. Hur & K. S. Choi. (2012). A Study on characteristics and types of tweet in twitter. Hanminjok Emunhakhoe, 61, 455-494.
  4. H. J. Kim. (2017. 07. 28). Twitter users remain stuck ... Stock price plummeted by 14%.. yonhapnews. http://goo.gl/3yjTD9
  5. M. W. Nho. (2012). Korea's Popular Celebrity Twitter Users and Celebrity Culture Cybercommunication Academic Society 29.4, 95-143.
  6. H. Y. Cho, H. J. Kim, E. C. Lee, M. J. Lee, Y. W. Nam & Y. H. Kim. (2017) Twitter Data Collectionto Build Customized Tweet Recommendation System, korea multimedia society,, 254-255
  7. Y. W. Nam & Y. H. Kim. (2016). A System of Storing Important Opinion about Twitter Trends, Korean Institution of Information Scientists and Engineering, 337-339.
  8. Y. W. Nam & Y. H. Kim. (2016). Improving Twitter Search Function Using Twitter API. Proceeding of journal of multimedia services convergent with art, humanities, and sociology 8, 879-886.
  9. S. J. Yang, J. W. Choi, S. H. Moon, Y. W. Jung, Y. W. Nam & Y. H. Kim. (2016). Opinion Mining Using Retweet Function of Twitter. Proceedings of Journal of The Korean Institute of Intelligent System, 26.1, 193-194.
  10. T. Mikolov, K. Chen, G. Corrado & J Dean. (2013). Efficient estimation of word representations in vector space. In Proceedings of Workshop at ICLR.
  11. T. Mikolov, I. Sutskever, K. Chen & GS. Corrado. (2013). Distributed representations of words and phrases and their compositionality. Advances in neural information processing systems. 3111-3119
  12. D. W. Ko & J. J. Yang. (2018). Korean Natural Language Processing and Analysis. Korean Institution of Information Scientists and Engineering. 2140-2142.
  13. D. W. Leem & H, Y, Jang. (2017). Keyword Extraction from Korean Wikipedia Using Word Similarity. Proceedings of Journal of The Korean Institute of Intelligent System, 850-852.
  14. E. L. Park & S. Z. Cho. (2014). KoNLPy: Korean natural language processing in Python. Proceedings of the 26th Annual Conference on Human & Cognitive Language Technology, 133-136
  15. E. L. Park & S. Z. Cho. (2014). KoNLPy: Python Korean NLP. goo.gl/1dPrka