To provide relevant data to users form massive data available on web the Semantic Web technique is used. This presentation gives introduction of semantic web and how NLP can be used in it.

1. “HANDLING UNSTRUCTURED
DATA FOR SEMANTIC WEB”
Subject : Introduction of Semantic Web using NLP
- Sandeep Wakchaure

2. CONTENTS
 Introduction
 Data Forms
 Architecture
 Example
 Conclusion
 Reference

3.  Problem Statement : To get knowledge or
information on web search engine which does not
contain any kind of irrelevant data.
 Reason : In day to day life there large amount of
unstructured data is going to stored on web, data
warehouses , repository or on cloud.
 Purpose of the System: To get the relevant
data there should be technique which process the
entered keywords, find the context and provide
the relevant knowledge.

4. INTRODUCTION
 When a user search on web i.e. retrieving data through search engine, the
results contains the large amount of data which are not user’s required
information.
 In this case keywords search techniques is fail here, to get required
information with the huge amount of information.
 Hence there is need to move from original Web to the Semantic Web for fast
related and precise information access.
 Many fields of computer world such as Data Mining, Information
Retrieval, Database Management System and NLP have been introduced
with Semantic Web for machine supported data interpretation and
process integration.

5. DATA IN THE FORM OF :
Structured Data-
This data has structure in terms of grammar pattern and
contextual relations.
Unstructured Data-
This data has not a specific structure but it may have grammar.
Posted queries and answers on the page, advertisements,
graphics, text, emails, presentations and so they are included in
the unstructured data.

6. TECHNIQUE USED
Ontology : Ontology is a description of things that exist and how they
relate to each other. It is a study of categories of things and their
relation among them.
The core part of Semantic Web is ontologies, which defines the
relationship between related entities, which achieved using
 RDF(S) (Resource Description Framework/Schema) and
 OWL (Web Ontology Languages)
Ontologies and reasoning rules are applied to reason about data and infer
new information. Rules are nothing but some condition or
restriction to be applied on data to draw some facts. In fact,
Semantic Web is like a collection of related and clustered facts.

7. For finding pattern from these sources, pre-processing of the source
documents required which is supported by the NLP techniques.
The techniques are like,
1.Stemming (finding stem)
2.Removing suffixes and prefixes
3.Lemmatization for replacing inflected word with its base form,
4. Part of Speech (POS) tagging for finding grammar category of
language - such as Noun, pronoun, adverb, adjective, proposition
Using ontologies with NLP, understanding of natural language through
systems become smarter enough to make inference and respond with
defined and relevant result what a user requests.
CONT…

8. SEMANTIC WEB LAYERED STACK

9. ARCHITECTURE

10. EXAMPLE :
Dependency graph for sentence : “on-screen keyboard
displays a virtual keyboard on computer-screen”

11. CONCLUSION
The goal of the system is to automate the software agents for the
retrieving relevant and required information or data rather
than providing the massive unrelated data.
NLP techniques with the Semantic web provide the capability to
turning the original web to Semantic Web while dealing with
a combination of structured and unstructured data.

12. REFERENCES
[1] Gharehchopogh FS, Khalifelu ZA ; Analysis and
Evaluation of Unstructured Data: Text Mining
versus Natural Language Processing. Application
of Information and Communication Technologies
(AICT), 5th International Conference, 2011
[2] https://en.wikipedia.org/
[3] Fortuna B, Grobelnik M, Mladenić D; OntoGen:
Semi-automatic Ontology Editor. Proceedings of
HCI, 2007;309-318
[4] https://www.quora.com/