Entity Linking based Graph Models for Wikipedia Relationships
Citation
Mattakoyya Aharonu , Mastan Rao Kale "Entity Linking based Graph Models for Wikipedia Relationships", International Journal of Engineering Trends and Technology (IJETT), V18(8),380-385 Dec 2014. ISSN:2231-5381. www.ijettjournal.org. published by seventh sense research group
Abstract
Measuring relationships between pairs of data objects in Wikipedia is challenging task in real world data. For the Wikipedia graph, consisting of the articles together with the hyperlinks between them, the preferential attachment rule explains portion of the constitution, but instinct says that the themes of each article also performs a crucial position. This proposed system concentrate on small datasets extracted from the Wikipedia database. The matter of researching individual search space intents has attracted intensive consideration from both enterprise and academia. However, state-of-the-art intent researching techniques go through from different drawbacks when only utilizing a unmarried variety of statistics supply. For instance, query textual content has issue in distinguishing ambiguous queries; search space log is bias for a order of seek outcome and users noisy click on behaviors. In this proposed system, we`ll use three kinds of similar objects, namely queries, websites and Wikipedia ideas collaboratively for getting to know generic search space intents and assemble a heterogeneous graph to characterize a number of kinds of relationships between them. A novel unsupervised system known as heterogeneous graph-based soft-clustering is developed to derive an intent indicator for each product depends on the constructed heterogeneous graph. Entity Linking (EL) is the duty of linking name mentions in Net textual content with their referent entities in a know-how base. Classic EL approaches generally hyperlink name mentions in a record by assuming them to be unbiased. However, there`s often additional interdependence between different EL judgements, i.e., the entities inside the same record ought to be semantically concerning one another. In these circumstances, Collective Entity Linking, wherein the name mentions within the same record are linked collectively by exploiting the interdependence between them, can increase the entity linking accuracy.
References
1. Wikipedia, the Free Encyclopedia. http://wikipedia.org, accessed in 2006.
2. David Aumueller. SHAWN: Structure Helps a Wiki Navigate. In Proceedings of BTW Workshop WebDB Meets IR, March 2005.
3. Francesco Bellomi and Roberto Bonato. Network Analisis for Wikipedia. In Proceedings of Wikimania 2005, The First International Wikimedia Conference. Wikimedia Foundation, 2005.
4. Abraham Bookstein, Vladimir Kulyukin, Timo Raita, and John Nicholson. Adapting Measures of Clumping Strength to Assess Term-Term Similarity. Journal of the American Society for Information Science and Technology, 54(7):611–620, 2003.
5. Sergey Brin and Lawrence Page. The Anatomy of a Large-scale Hypertextual Web Search Engine. Computer Networks and ISDN Systems, 30(1–7):107–117, 1998.
6. Ludovic Denoyer and Patrick Gallinari. The Wikipedia XML Corpus. Technical report, 2006.
7. Daniel Kinzler. WikiSense — Mining the Wiki. In Proceedings of Wikimania 2005, The First International Wikimedia Conference. Wikimedia Foundation, 2005.
8. Jon Kleinberg. Authoritative sources in a hyperlinked environment. Technical Report RJ 10076, IBM, 1997.
[10] V. Hristidis, L. Gravano, and Y. Papakonstantinou. Efficient IR-style keyword search over relational databases. In VLDB, pages 850–861, 2003.
[11] V. Hristidis and Y. Papakonstantinou. Discover: keyword search in relational databases. In VLDB, pages 670–681, 2002.
[12] K. Järvelin and J. Kekäläinen. IR evaluation methods for retrieval”