SPIRIT: A TREE KERNEL-BASED METHOD FOR TOPIC PERSON INTERACTION DETECTION
0 likes98 views
The document presents a method called SPIRIT for detecting topic person interactions in texts through a rich interactive tree structure that incorporates syntactic, context, and semantic information. Experimental results indicate that SPIRIT outperforms traditional relation extraction methods and offers a pathway for future research, particularly in enhancing interaction detection through sentiment analysis and keyword extraction. The method is adaptable to different languages, with plans to implement it for English texts using tools like the Stanford parser.
![CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249)
MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com
Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com
person interactions by adopting the English version of Stanford parser which
labels dependencies between English tokens, and replacing E-HowNet with an
English lexical database (e.g., WordNet) to calculate verb similarity. In the
future, we will also deploy SPIRIT for English texts.
REFERENCES
[1] R. Nallapati, A. Feng, F. Peng, and J. Allan, "Event Threading within News
Topics," Proc. 13th ACM International Conference on Information and
Knowledge Management, pp. 446-453, 2004.
[2] A. Feng and J. Allan, "Finding and Linking Incidents in News," Proc. 16th
ACM International Conference on Information and Knowledge Management,
pp. 821-830, 2007.
[3] C.C. Chen and M.C. Chen, "TSCAN: A Content Anatomy Approach to
Temporal Topic Summarization," IEEE Trans. Knowledge and Data Eng., vol. 24,
pp. 170-183, 2012.
[4] Y.C. Chang, P.H. Chuang, C.C. Chen, and W.L. Hsu,"FISER: An Effective
Method for Detecting Interactions between Topic Persons," Proc. 8th Asia
Information Retrieval Societies Conference, pp. 267-277, 2012.
[5] Y.C. Chang, C.C. Chen, and W.L. Hsu, "A Composite Kernel Approach for
Detecting Interactive Segments in Chinese Topic Documents," Proc. 9th Asia
Information Retrieval Societies Conference, pp. 215-226, 2013.](https://image.slidesharecdn.com/spiritatreekernel-basedmethodford-160722073145/85/SPIRIT-A-TREE-KERNEL-BASED-METHOD-FOR-TOPIC-PERSON-INTERACTION-DETECTION-3-320.jpg)
![CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249)
MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com
Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com
[6] M. Zhang, J. Zhang, J. Su, and G.D. Zhou, "A Composite Kernel to Extract
Relations between Entities with both Flat and Structured Features," Proc. 21st
International Conference on Computational Linguistics and the 44th Annual
Meeting of the Association for Computational Linguistics, pp. 825-832, 2006.
[7] L.H. Qian, and G.D Zhou, "TreeKernel-based Protein–Protein Interaction
Extraction from Biomedical Literature," Journal of Biomatical Informatics, vol.
45, no. 3, pp. 535-543, 2012.
[8] J. Xiao, J. Su, G.D. Zhou, and C.L. Tan, "Protein-Protein Interaction
Extraction: ASupervised Learning Approach," Proc. 1st International
Symposium on Semantic Mining in Biomedicine, pp. 51-59, 2005.
[9] G.M. Vernon, Human Interaction: An Introduction to Sociology. Ronald
Press Co., New York,1st edn., 1965.
[10] M. Collins and N. Duffy, "Convolution Kernels for Natural Language," Proc.
Annual Conference on Neural Information Processing Systems, pp. 625-632,
2001. [](https://image.slidesharecdn.com/spiritatreekernel-basedmethodford-160722073145/85/SPIRIT-A-TREE-KERNEL-BASED-METHOD-FOR-TOPIC-PERSON-INTERACTION-DETECTION-4-320.jpg)
SPIRIT: A TREE KERNEL-BASED METHOD FOR TOPIC PERSON INTERACTION DETECTION
- 1. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com SPIRIT: A TREE KERNEL-BASED METHOD FOR TOPIC PERSON INTERACTION DETECTION ABSTRACT: The development of a topic in a set of topic documents is constituted by a series of person interactions at a specific time and place. Knowing the interactions of the persons mentioned in these documents is helpful for readers to better comprehend the documents. In this paper, we propose a topic person interaction detection method called SPIRIT, which classifies the text segments in a set of topic documents that convey person interactions. We design the rich interactive tree structure to represent syntactic, context, and semantic information of text, and this structure is incorporated into a tree- based convolution kernel to identify interactive segments. Experiment results based on real world topics demonstrate that the proposed rich interactive tree structure effectively detects the topic person interactions and that our method outperforms many well-known relation extraction and protein-protein interaction methods. CONCLUSION A topic is associated with specific times, places, and persons. Knowing the interactions between the topic persons can help readers construct the background of the topic and facilitate document comprehension. In this paper, we have proposed a topic person interaction detection method called SPIRIT, which identifies text segments mentioning person interactions in a set of topic
- 2. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com documents. We designed a rich interactive tree structure which employs three RIT operators to explore the syntactic and semantic information in a text. The information is then applied to the convolution tree kernel to classify the interactive segments. The experiment results based on real-world datasets demonstrate that the proposed RIT operators can successfully exploit the syntactic structures, interaction semantics, and segment context relevant to person interactions. Consequently, our method outperforms many well-known relation extraction and tree kernel-based PPI methods. Current results have paved the way for other potential research topics. For instance, we observed that person interactions generally involve sentiments. The sentiment information of a text can be investigated to enhance our rich interactive tree structure and to improve the interaction detection results. Furthermore, developing effective interaction keyword extraction methods would be worthwhile in order to better represent person interactions. In our performance evaluation, CK’s precision shows the potential of content similarity for interaction detection. Instead of compositing another content kernel (e.g., the bigram kernel), we are designing a new RIT ornamenting operator in which tokens in the person dependency path will be examined to measure the content similarity between RITs. Moreover, concepts of the tokens will be incorporated to increase both detection precision and recall. We will also employ techniques of word sense disambiguation to improve the verb similarity calculation of RIT ornamenting. Finally, SPIRIT is adaptable to different languages. Taking English as an example, we can detect English
- 3. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com person interactions by adopting the English version of Stanford parser which labels dependencies between English tokens, and replacing E-HowNet with an English lexical database (e.g., WordNet) to calculate verb similarity. In the future, we will also deploy SPIRIT for English texts. REFERENCES [1] R. Nallapati, A. Feng, F. Peng, and J. Allan, "Event Threading within News Topics," Proc. 13th ACM International Conference on Information and Knowledge Management, pp. 446-453, 2004. [2] A. Feng and J. Allan, "Finding and Linking Incidents in News," Proc. 16th ACM International Conference on Information and Knowledge Management, pp. 821-830, 2007. [3] C.C. Chen and M.C. Chen, "TSCAN: A Content Anatomy Approach to Temporal Topic Summarization," IEEE Trans. Knowledge and Data Eng., vol. 24, pp. 170-183, 2012. [4] Y.C. Chang, P.H. Chuang, C.C. Chen, and W.L. Hsu,"FISER: An Effective Method for Detecting Interactions between Topic Persons," Proc. 8th Asia Information Retrieval Societies Conference, pp. 267-277, 2012. [5] Y.C. Chang, C.C. Chen, and W.L. Hsu, "A Composite Kernel Approach for Detecting Interactive Segments in Chinese Topic Documents," Proc. 9th Asia Information Retrieval Societies Conference, pp. 215-226, 2013.
- 4. CONTACT: PRAVEEN KUMAR. L (, +91 – 9791938249) MAIL ID: sunsid1989@gmail.com, praveen@nexgenproject.com Web: www.nexgenproject.com, www.finalyear-ieeeprojects.com [6] M. Zhang, J. Zhang, J. Su, and G.D. Zhou, "A Composite Kernel to Extract Relations between Entities with both Flat and Structured Features," Proc. 21st International Conference on Computational Linguistics and the 44th Annual Meeting of the Association for Computational Linguistics, pp. 825-832, 2006. [7] L.H. Qian, and G.D Zhou, "TreeKernel-based Protein–Protein Interaction Extraction from Biomedical Literature," Journal of Biomatical Informatics, vol. 45, no. 3, pp. 535-543, 2012. [8] J. Xiao, J. Su, G.D. Zhou, and C.L. Tan, "Protein-Protein Interaction Extraction: ASupervised Learning Approach," Proc. 1st International Symposium on Semantic Mining in Biomedicine, pp. 51-59, 2005. [9] G.M. Vernon, Human Interaction: An Introduction to Sociology. Ronald Press Co., New York,1st edn., 1965. [10] M. Collins and N. Duffy, "Convolution Kernels for Natural Language," Proc. Annual Conference on Neural Information Processing Systems, pp. 625-632, 2001. [