International Journal of Engineering and Technical Research (IJETR)

ISSN: 2321-0869, Volume-3, Issue-6, June 2015

An Advance Way to Retrieve Information from Web

by Natural Language Interface
Dr. Mayura Mathankar

assistance to a user dealing with a particular computer

Abstract Information is one of the most important part of application. Interface agent must have some kind of
our todays daily life. There are large numbers of things that can intelligence, such as knowledge acquisition, autonomy and
be done using the information and so it is very important. We collaboration. The architecture of intelligent interface
can say that with the progress in the information technology, we includes rules, frame descriptors, discrimination networks,
are progressing in every sphere of life as it not only makes our inference engine, associative memory, matching and
tasks easier but also saves a lot of time. To access any autonomous agents.
information, internet is used for different purposes depending
upon the requirement. People mostly rely on internet for Intelligent user interfaces have been proposed as a means to
acquiring desired information. Search engine plays vital role in overcome some of the problems that direct manipulation
retrieving the required information in fast way. interfaces cannot handle, such as information overflow
Many existing state-of-the-art search engines face challenges problems; providing help as how to use complex systems; or
of effective information retrieval. This research paper deals real time problems. Intelligent user interface is also being
with the design of Interface for Web-based information proposed as a means to make systems individualized or
Retrieval with the help of Natural Language. In this proposed personalized. Interface agents are the prevailing development
model, natural language query is given to the developed search towards intelligent user interfaces.
engine to understand its meaning and depending on its meaning
the most relevant documents are retrieved from the World Wide
An intelligent interface cooperates with the user in
performing its tasks, working as a personal user assistant. The
Web.
interface agent is pro-active taking the initiative and not
The proposed algorithm is based on Statistical Machine
passive. Usually, an intelligent interface is not the interface
Learning. Natural Language text is converted to more formal
between the user and the application. Instead, it observes the
representation such as first-order logic structure because it is
interactions between the user and the program learns with it
easier for computer programs to manipulate. The aim of this
and interacts both with user and program.
research work is to provide interactive interface to the search
engine to enhance its effectiveness for information retrieval.
1.2 Web-based information retrieval
Index TermsNatural Language, Algorithm, Web-based The Internet and the Web offer new opportunities and
information. challenges to information retrieval researchers. With the
information explosion and never ending increase of web
pages as well as digital data, it is very hard to retrieval useful
I. INTRODUCTION and reliable information from the Web. Materials from
From computers and smartphones to watches and eyewear, millions of web pages from organizations, institutions and
almost anything can now be connected to what is essentially personnel have been made public electronically accessible to
one huge global network, which is constantly evolving millions of interested users. The Web uses an addressing
further. The internet has revolutionised communication and system called Uniform Resource Locators (URLs) to
networking to the point where few people in the world cannot represent links to documents on web servers. These URLs
be reached in a few seconds at the click of a button. Gone are provide location information. Like titles of books in
the days of sending letters and waiting days for a reply, being traditional libraries, no one can remember all URLs on the
the only person in a specific area with a specific interest or the Web. Web search engines allow us to locate the Internet
inability to speak to people face-to-face without being in resources through thousands of Web pages. It is almost
proximity of them, replaced by the instant contact method of impossible to get the right information as there is too much
email. People are increasingly accustomed to have resource to irrelevant and out dated information.
computer programs to do the computations, to search needed
information and even to make decisions. Therefore, an Information retrieval systems provide useful information.
obliging and clever interface will greatly reduce users work The Web can be viewed as a virtual library. Information
and increase the efficiency of the computer programs. retrieval is an important and major component of the Internet.
Most of the current search engines are based on words, not the
1.1 Intelligent Interfaces concepts. When searching for certain information or
Intelligent interface agents are computer programs that knowledge with a search engine, one can only uses a few key
employ artificial intelligence techniques in order to provide words to narrow down the search. The result of the search is
tens or maybe hundreds of relevant and irrelevant links to
Dr. Mayura Mathankar, Associate Professor, Sydenham College of
various Web pages.
Commerce & Economics, Mumbai

137 www.erpublication.org
 An Advance Way to Retrieve Information from Web by Natural Language Interface

Document representation, query formulation, and retrieval

functions are fundamental issues in information retrieval Objects: people, city, numbers, colors, football etc.
study. Based on this classification, we have to design and Relations: red, round, prime, brother of, bigger than, part
implement an appropriate scheme to represent the contents of of etc
documents, a language to express user queries, and a retrieval Functions: father of, best friend, one more than, plus etc
function to search for relevant documents. Index terms play
the connecting role between documents and user queries. A The syntax of FOPL consists of following elements
document is considered to be relevant to a query if the user
submitting the query judges the document to be useful. ConstantsKrishna, 2, NUS,...
Different retrieval models have been developed, such as the PredicatesBrother, >,...
Boolean, vector space, and probabilistic models, as well as FunctionsSqrt, Left, Upper
recent linguistic and knowledge-based models. The first three Variablesx, y, a, b,...
models are often referred to as the exact match model; the Connectives , , , ,
latter as the best match models. Although they provide us
formal and elegant formulations of information retrieval 3 Rules
problems, they suffer from several shortcomings. The Complex deductive arguments can be judged valid or invalid
classical retrieval models are over-simplification of the real based on whether or not the steps in that argument follow the
world retrieval problem. basic rules of inference. These rules of inference are all
relatively simple, although when presented in formal terms
II. METHODOLOGY they can look overly complex.

A natural language search engine would find targeted answers 4 Inference Engine
to questions. For example, when confronted with a question of An inference engine is a software system that is designed to
the form which Indian state has the highest income tax?', draw conclusions by analysing problems in light of a database
conventional search engines ignore the question and instead of expert knowledge it draws upon. It reaches logical
search on the keywords 'state, income and tax'. Natural outcomes based on the premises the data establishes.
language search, on the other hand, attempts to use natural Sometimes inference engines are also capable of going
language processing to understand the nature of the question beyond strict logical processing, and utilize probability
and then to search and return a subset of the web that contains calculations to reach conclusions that the knowledge database
the answer to the question. If it works, results would have a doesn't strictly support, but instead merely implies or hints at.
higher relevance than results from a keyword search engine.
In interface design natural language interfaces are sought after 5 Semantic Parser
for their speed and ease of use, but most suffer the challenges
to understanding wide varieties of ambiguous input. Semantic parser checks syntax of the text, i.e. whether the text
is grammatically correct or not. It further analyses the input
It is important to note that text interfaces are 'natural' to text for understanding its meaning. The meaning of any
varying degrees, and that many formal (un-natural) natural language sentence depends on the way of parsing the
programming languages incorporate idioms of natural human text and its context. For example, consider a sentence, When
language. Likewise, a traditional keyword search engine I sing well children feel sick. This sentence implies two
could be described as a 'shallow' Natural language user meanings based on how we parse it.
interface. If search engine is supplemented by an intelligent
interface, it may help users to get relevant information Ex.1 : When I sing well, children feel sick
effectively and efficiently which can result in optimum
utilization of e-resources by potential users. Ex. 2: When I sing, well children feel sick

III. INTELLIGENT NATURAL LANGUAGE 6 Search Engine

INTERFACE
A web search engine is designed to search for information on
Many intelligent interfaces exhibit mainly three features such
the World Wide Web. The search results are generally
ie Knowledge Acquisition, Autonomy and Collaboration. The
presented in a line of results often referred to as search engine
Intelligent Natural Language Interface for Web based
results pages (SERPs). The information may be a specialist in
information retrieval is comprised of following components:
web pages, images, information and other types of files. Some
search engines also mine data available in databases or open
1 Knowledge Acquisition :
directories. Unlike web directories, which are maintained
It is the process of absorbing and storing new information in
only by human editors, search engines also maintain real-time
memory, the success of which is often gauged by how well the
information by running an algorithm on a web crawler.
information can later be remembered.

2 Knowledge representation
For knowledge representation First- Order Predicate Logic
will be used. First-order logic (like natural language) assumes
the world contains:

138 www.erpublication.org
 International Journal of Engineering and Technical Research (IJETR)
ISSN: 2321-0869, Volume-3, Issue-6, June 2015

Highlights Nouns &Adjectives

Nouns &Adjectives Highlighted
Intelligent
Natural
User Image Extraction Language Interface

Image Extracted

Java keywords highlighting

Java keywords highlighted

Figure 1: DFD for Intelligent Natural Language Interface

IV. RESULTS
The proposed system is developed and implemented in JAVA.
Rigorous experimentation was carried out. The results are REFERENCES
tested for various sample queries.
[1] Han, H., Extracting news from server side databases by query
Precision and recall interfaces (2014) Journal of Computer Information Systems, 54 (2),
Average Average Recall pp. 57-65
[2] Cobos, C., Muoz-Collazos, H., Urbano-Muoz, R., Mendoza, M.,
Precision Len, E., Herrera-Viedma, E., Clustering of web search results based
Without 0.64 0.72 on the cuckoo search algorithm and Balanced Bayesian Information
interface Criterion (2014) Information Sciences, 281, pp. 248-264
[3] Sleiman, H.A., Corchuelo, R., A class of neural-network-based
With interface 0.81 0.76 transducers for web information extraction (2014) Neurocomputing,
135, pp. 61-68.
[4] A.K.Rabiah, T.M.T. Sembok, B.Z. Halimah, Improvement of
document understanding ability through the notion of answer literal
V. CONCLUSION expansion in logical-linguistic approach, WSEAS Transactions on
In this paper we have presented the design of Natural Information Science and Applications, Vol. 6, Issue 6, June 2009
[5] Marco Aiello, ChristofMonz, Leon Tororan, Marcel Worring,
Language Interface for Web-based information Retrieval. In
Document Understanding for a Broad Class of Documents,
this model, natural language query is analysed for International Journal on Document Analysis and Recognition.
understanding its meaning and depending on its meaning the [6] Javier Albusac, David Vallejo, and J.J. Castro-Schez, Paolo Remagnino,
most relevant documents are retrieved from the World Wide Carlos Glez Morcillo and Luis Jimenez, Monitoring Complex
Web. Digital document libraries have become an Environments Using a Knowledge- Driven Approach Based on
Intelligent Agents, IEEE Journal of Intelligent Systems,May- June
increasingly important means of storing information within 2010, Vol. 25, no. 3, pp : 24 - 31
organizations. Automatic understanding of documents will
certainly improve the utilization of e-resources among the
potential users.The aim of this work is provide interactive
interface to the search engine to enhance its effectiveness for
information retrieval.

139 www.erpublication.org