The Content Search By Using Semantic Web

Published Date: 02 Nov 2017

using Semantic Web

Abstract

The content search of web content for large data and information presents enormous resourcing and quality challenges. Users expect to find information quickly, with minimal navigation and with consistency of information and nomenclature. For example content and solutions information, users expect clear, relevant lists of information and services that comprise those solutions, including research papers, publications, videos, images, interviews, conferences and case studies that provide referential examples. The foundation component of the Semantic Web is ontologies which are used to embody knowledge in the Semantic Web. Ontology is a data model which can be used to illustrate a set of concepts and the relationships connecting those concepts within a domain. Taking unstructured data from the web and formalizing it so that it can be structured automatically is a difficult work to do, but apart from its significance it is interesting as well. Through this research it is intended to make the automation of ontology public, as it is based on open standard and constructed using publicly available resources of Google, like Google AJAX API and JavaScript parser JSON.

Introduction

Today, people of the world are dependent on World Wide Web search engine, and changes on these search engines can be seen weekly. Those of us who have to get real work done using these engines just want to know which ones we should use when, and what we should know about how they work [1].

People of the world are dependent on worldwide web. A proper search engine consists of database and tools which generate database .Most search engine allow you to type your query and find out appropriate result for you. But some of us are not satisfy by that results, because sometimes that result is not related to query we asked. Semantic web consist of Syntax means "meaning behind the sentence". But still semantic web content search is not enhance .In our article we have introduce Enhance Semantic search ESS technique to enhance the content of Semantic web which improve quality and consistency of the web.

A search engine proper is a database and the tools to generate that database and search it; a catalog is an organizational method and related database plus the tools for generating it. There are sites out there, however, that try to be a complete front end for the Internet [1].

Most engines allow you to type in a few words, and then search amount of these words in their data base. Each engine has their own way of deciding what to do about approximate spellings, plural variations, and truncation. If you just type words into the "basic search" interface you get from the search engine's main page, you also can get different logical expressions binding the different words together [1].

Most engines have separate advanced search forms where you can be more specific, and form complex Boolean searches .Some search tools parse HTML tags, allowing you to look for things specifically as links, or as a title or URL without consideration of the text on the page [1].

The Semantic Web Introduction

The Semantic Web is an addition of the World Wide Web with new technologies and values that enable processing of data and useful information for mining by a computer. The Semantic Web provides a understanding technique to fill the gap between human being and computers. The Semantic Web is an addition of the current Web with new techniques and values that enable to understand different type of software applications. The Web controls a huge amount of data but the computers alone cannot recognize or make any decisions with this data. The solution for this problem is the Semantic Web. The Semantic Web is not a split Web but an addition of the current one, in which information is given well defined sense and meaning [2].

Ontologies in Semantic Web

The foundation component of the Semantic Web is ontologies which are used to embody knowledge in the Semantic Web. Ontology is a data model which can be used to illustrate a set of concepts and the relationships connecting those concepts within a domain. OWL endows with three increasingly expressive sublanguages: OWL Full, OWL DL and OWL Lite [2]. Presented Semantic Web ontologies can be categorized into the following four major categories: comprehensive, meta-ontologies, systematic domain specific ontologies, upper ontologies and simple specialized ontologies [3].

Searching on the Semantic Web

There are several Semantic Web search engines e.g. Kosmix, Swoogle, Factbites, Exalead, Power set, Sensebot etc, they are used to explore and retrieve the ontologies from the Web but most of the search engines are not yet put into practiced since they are in research level. The Semantic Web search engine can be roughly divided into two categories; one is Ontology search engine and other is Semantic search engines. Our effort is to find out Ontology search engines which are exercised to find the Semantic Web documents; they use a technique named ranking technique to find out the closest results of the query which user asked for search [3].

Internal Structure (IS) and Semantic Web Link Structure (SWLS) Based Ontology Ranking Technique.

This technique is relied mainly on both internal structure (IS) and Semantic Web link structure (SWLS) of the ontology. They assign more weighting for internal structure (IS) and low weight for Semantic web link structure (SWLS) to get the most excellent ranking. Semantic Web Link structure (SWLS) checks the popularity of the ontology and internal structure (IS) considers how ontology characterizes the concepts and their relationships among themselves. This new ranking technique is relied on the criteria that there is involvement from the internal structure (IS) and Semantic Web link structure (SWLS) for the ranking of chosen ontologies [3].

Content-based Ontology ranking based method:

Content-based Ontology Ranking established a ranking method that is focused on the content likeness of an ontology to a corpus that is brought into consideration for the given search terms. In order to rank the ontologies, our ESS system will attempt to find a corpus that narrates to the domain that the user requires ontology to symbolize [3].

Semantic Relatedness

In this paper we introduce a solution to the very common problem with Ranking pages on the web and finding out the weight of the page to estimate the relevance of documents with respect to query which is being asked by user. Enhance Semantic search (ESS) ability are needed to overcome the limitations of long-established search engines that are mainly keyword based search engines. Ontologies play main role to attain this goal. Ontology is defined as "ontology defines as an explicit and formal specification of shared conceptualization". In our work ontology can be used as set of terms and relationship used between different concepts [1].

Problem Statement:

"To Enhance Content Search of semantic Web by using Google API to improve Content search Quality and Consistency based on ontology ranking method."

Problem Domain

This vision of a semantic Web is extremely motivated and would require solving many long-standing research problems in knowledge representation and reasoning, databases, computational linguistics, computer vision, and agent systems [4]. One such problem is the trade-off between conflicting requirements for expressive power in the language used for semantic annotations and the scalability of the systems used to process them; another is that Integrating different ontologies may prove to be at least as difficult as integrating the resources they describe. Emerging problems include how to create suitable annotations and ontologies and how to deal with the variable quality of Web content. [4]

The content search of web content for large data and information presents enormous resourcing and quality challenges. Users expect to find information quickly, with minimal navigation and with consistency of information and nomenclature. For example content and solutions information, users expect clear, relevant lists of information and services that comprise those solutions, including research papers, publications, videos, images, interviews, conferences and case studies that provide referential examples.

Automated services will improve in their capacity to assist humans in achieving their goals by "understanding" more of the content on the web, and thus providing more accurate filtering, categorization, and searches of information sources. This process will ultimately lead to an extremely knowledgeable system that features various specialized reasoning services. These services will support us in nearly all aspects of our daily life making access to information as pervasive, and necessary, as access to electricity is today.

Applications based on semantic technologies offer new ways to discover, browse and explore information. But how can we (as a semantic web "insider") explain these potential benefits to a typical end-user, who has never heard anything about "faceted search" and Enhance the context of Semantic Web by making Semantic web search Engine more Intelligent and User Friendly.[6]

Background

The Semantic Web augments the current WWW by giving information a well defined meaning, better enabling computers and people to work in cooperation. This is done by adding machine understandable content to Web resources. Such added content is called metadata, whose semantics is provided by referring to ontology—a domain’s conceptualization agreed upon by a community. [7]

Semantic Web was introduced by Tim burners Lee. Semantic Web is related to the "Syntax "how you declare something. Semantic web is the Meaning behind what you Say? Today the web is turning into Semantic Web, even Google support Micro format and RDF in documents and use metadata to make search results more relevant. [4]

According to Tim Berners-Lee,

'The Semantic Web will bring structure to the meaningful content of Web pages, creating an environment where software agents roaming from page to page can readily carry out sophisticated tasks for users.'

Objective of Study

Today’s Web is a relatively simple artifact. Web content consist of Hypertext and Hypermedia and it is simply Accessible via Link Navigation, One of the Web Strength is Simplicity even naive users quickly learn to use it and even create their own content. The explosion in both the range and quantity of Web content also highlights serious shortcomings in the hypertext paradigm. The required content becomes increasingly difficult to locate via search and browse. The Objective of this Research is to enhance the content of Web Search by creating suitable annotations and ontologies and how to deal with the variable quality of Web content.

Limitations of the Study:

The study that was conducted has the following limitations that need to be overcome in order to get better results.

To identify keyword and sub keyword for ontology construction user must have domain knowledge and identification of keywords is manual.

Literature Review

Our methodology for this Research Article is Empiricism which arises from sense experience, which is the part epistemology means theory of knowledge, what is knowledge and how it is acquired, what do people know about it. First we will observe the construction method of ontologies for specific domains, and then we will see the results of Content search of semantic web and we will introduce a technique which will enhance the content search of Semantic web.

Chapter 1 – Semantic Web

Introduction

The internet is one of the biggest inventions in the field of computers. In the start it was used as the normal web or web of information and data but with the enhancement of web through the concept of "semantic web" the human can interact with the web [1]. Semantic Web is the web which can be handled by the software (website) with the interaction of end user to make the application more efficient [2]. Search Engines are one of the categories of Semantic Web. We have structured and unstructured data on web like different file extensions, research data from different domains, etc. For tis we need to understand a concept of ‘Ontology’ and different languages that supports this concept and is being discussed in the literature Review.

Languages and Techniques

There are many languages and techniques for web but here we are interested in core concept as they are the building blocks of web.

Resource Description Framework:

RDF is one of the language which is used on the web and is used to assign a unique Universal Resource Identifier (URI) for a particular node [1].In RDF we create a node then link this node to its representative node to make a relational object as to move the web from the document to data Let me elaborate this with the help of diagram. In Figure (1) [1] there is a person Eric Miller, then we shows a kind of it i.e. person and then contact me but contact me at which address so we linked another node mail to, together they make a relationship and interact together.

Figure 1: Shows an RDF graph representing of Eric miller

OWL (Web Ontology Language):

OWL is the extension of RDF or XML document [5]. The basic idea of OWL is to identify weather the particular Ontology logically falls in the category of Ontology, it uses the linking provided by RDF document [1] and make its representation more efficiently and now it becomes a standard for semantic web. There are three types of OWL based on its uses in applications [6].

Folksonomies and Semantic Web:

Occurs when majority of peoples are interested in a particular information or want to tag it or tag their own content for retrieval[1], this is done by using keywords in the document of your web. There are number of application like flickr or sharing content sites, with the time passes the web is being transfer from the simple web to Semantic web.

Web Mining:

Refers to a concept how two entities are link together on the web or to find the strength of their relationship. We do this to reduce the no of query to search the particular record or name. Consider any two names at random like justin and smith, then put an edge between them and suppose some ratio of there co-occurrence, if both the names occurs on the web having an overlapping ratio we suppose then it means they have a strong relationship [3], but keep in mind that if the name is common like john then we must have to add some extra word for the sack of name-disambiguation [7].

Ontology from different perspective

There are many languages and techniques for web but here we are interested in core concept as they are the building blocks of web.

Ontology emergence in del.icio.us:

According to the author Joshua Schachter del.icio.us is a social networking tool for the person of book marking [4]. It’s a web based system where user can manage their favorite links and websites and describe them with the keywords or by tagging. User can post-bookmark related to the current seniors like politics, business, media etc and it’s a faster process as we considered it on the blogs. It’s an Actor-Concept-Instance [4]. The Tagging data is exposed in the form of RSS feed which is shown on the browser.

Community-based ontology extraction from web pages:

This is the extension of the Actor-Concept-Instance [4]. In this scenario we talked about community based book marking or tagging. Suppose we have a community whose members are actors for us and we have some terms randomly as concept and instances are the web pages in our model. Assume that the web page is tagged if concept (term) occurs on it. We have a search engine which counts the concept occurs on the page and by querying it we can get the co-occurrence of them and find the association of them as we do it Folksonomies [1], with the higher association we can get the community based tagging or ontology extraction from the web-pages.

Ontology and Folksonomies:

The point of debate here is that Ontology and Folksonomies falls in the same category? If no then is there a place of Folksonomies in the Semantic Web? [4]. The Author says that there is no need to make such a choice, in fact folksonomies are ontologies. But some people have different opinion [4]. Let’s compare these two: Ontology has a hierarchy with the subclasses relation and are lightweight as compared to folksonomies. Folksonomies are much less volatile because tagging content may changes on daily basis but investigating the folksonomies is the future work to make it more efficient [9].

Chapter 2 – Different Research Papers

Semantic-Powered Research Profiling

This paper is written by Zhixiong Zhang (Director of Information System department at National Science Library), Ying Ding Ph. D (Assistant Professor of Information Science Core Faculty of Cognitive Science), Na Hong (National Science Library Chinese Academy of Sciences); this paper was published at 8th international conference of semantic web 2009. This paper describes a novel infrastructure to generate semantic-powered research profiling for research fields, organizations and individuals. It crawls related websites and news feeds, extracts research terms, research objects and relations from them and uses the proposed Research Ontology to model them into RDF triples to facilitate semantic queries and semantic mining on burst detection, hot topic detection, dynamics of research, and relation mining.

Semantic Enhancement for Enterprise Data Management

This paper is written by Li Ma, Xingzhi Sun, Feng Cao, Chen Wang, Xiaoyuan Wang, Nick Kanellos,Dan Wolfson, Yue Pan, published at 8th international conference of semantic web 2009.In this paper researchers describes Semantic web technologies by taking customer data as an example, the paper presents an approach to enhance the management of enterprise data by using Semantic Web technologies. Customer data is the most important kind of core business entity a company uses repeatedly across many business processes and systems, and customer data management (CDM) is becoming critical for enterprises because it keeps a single, complete and accurate record of customers across the enterprise.

Semantic Web Search Ontology Ranking Algorithm

Semantic Web Search and Ontology Ranking Algorithm this article was written by Rajapaksha and N. Kodagoda ; Department of Information Technology; Sri Lanka Institute of Information Technology, Colombo . In this article they started with the introduction of semantic web and they describe about different semantic web search and ranking techniques in semantic web [2].

In this article they discussed about OntoSearch which uses Google API to get first 100 Ontologies. Content-based Ontology Ranking in which a ranking method is used to focus on the content similarity of an Ontology and ReConRank this method uses the well-known Page Rank/HITS algorithms to rank the Semantic Web data. This method combines ranks from the RDF graph with ranks from the context graph, i.e. data sources and their linkage [2].

An ontology-driven approach for Semantic Information Retrieval on the Web

Antonio M. Rinaldi, he describes Semantic relatedness which is used to evaluate the relevance of documents on the web with respect to query which is asked by user. In his DYse approach he considered three major techniques, which are metrics, semantic relatedness and WordNet. Metrics are applied on Semantic relatedness to find out the relationship between two words. WordNet is used for dictionary as well as to measure the semantic relatedness of two nouns related to each other and also find distance between them by keeping subject keyword and domain word. This approach is applied on Dynamic semantic network (DSN) [1].

Interactive Search using Google API and JSON

Different users uses the search engine by different ways e.g. user A might write ‘C++ Programming’’ and user B might write ’Programming C++’, the existing search engine gives the different output for both the Queries but logically its wrong, we need to manage this type of things so the logic here is that search must be refined by the keyword, the url must be selected for the combinational Search. Google API allow us to manage such things. The proposed solution in this paper is that an interface should be build which takes the keyword from user and then it connects to the GoogleAPI for the URL, The URL is being parse to the natural language by using JSON, for human understanding. Then these links will be filtered according to the combination of keywords.

Methodology

Architecture

The methodology used in exploring URLs for constructing the final ontology along with the selection of the class names is describe here.

Figure 3 demonstrates the detailed mechanism for analysis of websites in huge amount, so that the domain can find its related concepts by going through the inter-related keywords. To find the output with most accuracy JSON is used for its processing. Further to it, final ontology is constructed using 2 elements: first is, classes: selected concepts, and second is, OWL: language. Each concept is associated with the URL where the concept is extracted from. To present the most feasible ontology hierarchical order, the said process is run recursively to achieve the appropriate results.

Figure 3: Main Architecture

The following sections explain the process in detail.

Google AJAX API Search

Figure 4 depicts the architecture for interaction between Google Web Services and front-end application. Application developers have choice to use any programming or scripting language (Java, .Net, php, Python, Perl) they are comfortable with to build the connection with Google Web APIs services available remotely. Users’ queries for searching/extracting any information are processed in Google Server. And all of the above communication is done through Google AJAX Search API, as there is very less coding involved in integration on web page of Google’ search mechanism and its controls.

These include[10]:

a) Web Search: This is common as everyone uses it for searching the desired information from web, they simply enter their queries, and they get a list of search results on web page.

b) Local Search: The searching here is performed on specific location using Google Map.

c) Video Search: Video search results are extracted using the AJAX Video Search.As soon as it is builds its connection, the application will process search requests, and those requests are after processing in Google’s index and then spell check is performed in Google cache, the structured and accessible information is produced.

The basic functionality of the AJAX APIs is the integration of the hosted services with customized web pages, it allows this through JavaScript code, and for this reason Google’s widely known hosted services like Google Search and Google Maps are enhanced, and can be directly accessed by anyone.

Figure 4: Architecture of Google Search API

The core JavaScript code’s methods for searching is Search.execute ( ) and for feeding is Feed.load ( ). Once the request is received at Google server, the above mentioned methods are executed and response is generated on web page either using JSON or XML formats. On the other hand, the parsing is done in either way manually or automatically through provided UI controls of AJAX APIs.

JSON

JavaScript Object Notation (JSON) as shown in Figure 3 is designed for making data easily readable for humans without creating any heavy process as it is text-based open standard. Its derivation is as its name indicates from the JavaScript programming language to make data structures and objects simpler. Though it is associated with JavaScript, but still it has parsers available thus making it language independent (i.e. any programming language can utilize it). Figure also shows on of these forms that the String data structure can take. JSON Schema defines the structure of JSON data, and how it can be utilized in particular application and how it can get modified accordingly, basically it is specification for JSON-based format. Its concept is taken from XML Schema which is used for XML format, and provides features such as documenting (self descriptive), validating, and interacting JSON data [17].

Figure 5: JSON Schema

The system we are proposing through this research has used JSON to great extent in parsing the Google API response, as it gets less complex, and through JSON system can store the results in array, then process with retrieving the desired URLs, its content, and count result. It has also helped in the analysis of the candidate words [17].

Ontology Representation

Ontology is basically representation of the knowledge segregated in sets, so the concepts of same domain can be easily understood [3], and it is the vocabulary extension of Resource Development Framework (RDF). Knowledge and concepts are inter-related with each other, for acquiring the knowledge we need to clarify the concepts, and for clarification we need to understand the actual context of data. Ontology is discovered for the same purpose, so that web search can become easier for everyone. Ontology is a data model which can be used to illustrate a set of concepts and the relationships connecting those concepts within a domain. OWL endows with three increasingly expressive sublanguages: OWL Full, OWL DL and OWL Lite [5]. Presented Semantic Web ontologies can be categorized into the following four major categories: comprehensive, meta-ontologies, systematic domain specific ontologies, upper ontologies and simple specialized ontologies [6].

Figure 6: Social Network Ontolgy on Protégé 3.4.1

Protégé has plug-in called Jambalaya [21] that provides user with graphical presentation of the visualized hierarchy (Jambalaya extension can be easily installed with Protégé). Jambalaya uses Shrimp for visualization of Protégé -Frames and Protégé –OWL ontologies. Refer Figure 5 for ontology with URLs for the keyword Social Network.

Figure 7: Jambalaya tab for Network Ontolgy on Protégé 3.4.1

Design and Implementation

Following procedure has been followed for the implementation of the application using Java and JSP. The detailed architecture is been discussed in context of Figure 3.1.

1. The process begins with choosing a keyword Social Network, for instance and then entering it keyword in the HTML interface, for which the ontology is basically constructed;

2. Once we provide the input in terms of keyword, the program uses Google AJAX Search API for output i.e. to retrieve the information of that keyword along with the URLs that contains that keyword.

The following code is used to connect to Google,

query = URLEncoder.encode(query, "UTF-8");

URL url = new

URL("http://ajax.googleapis.com/ajax/servi

ces/search/web?start=0&rsz=large&v=1.0&q="

+ query);

// opening connection

URLConnection connection =

url.openConnection();

3. The result for our entered keyword Social Network is is shown in Figure 6. The provided result (data) is in the form of an array that includes all useful information of the matching websites, titles, URLs, etc. The response date is mentioned below, filtering is applied over it, for example, putting restriction on the number of returned results (100 is the number in our case), and many other filters are used in the program.

4. For the construction of the ontology, it is necessary to parse the data, so that class and URL selection becomes more appropriate. JSON has been used for parsing the response data. Parsing makes it easier for the program to split the websites retrieved against the keyword provided and also capture relevant results for the keyword (Social Network).

Following code is used for parsing the provided result data using JSON:

// Get the JSON response

String line;

StringBuilder builder = new StringBuilder();

BufferedReader reader = new BufferedReader(new

InputStreamReader(connection.getInputStream()));

while((line = reader.readLine()) != null) {

builder.append(line);

out.println("\n\n"+line);

}

String response1 = builder.toString();

//Parsing of response data

out.println("Total results =

"+json.getJSONObject("responseData").getJSONObject("c

ursor").getString("estimatedResultCount")+"<br>");

out.println("<br>");

JSONArray ja =

json.getJSONObject("responseData").getJSONArray("resu

lts");

out.println("Results:<br>");

for (int i = 0; i < ja.length(); i++) {

out.println("<br>");

JSONObject j = ja.getJSONObject(i);

out.println("Title:"+j.getString("titleNoFormatting"));

out.println("<br>");

out.println("URL:"+j.getString("url"));

out.println("<br>");

out.println("Content:"+j.getString("content"));

5. URLs selection for the class is dependent upon the occurrence of the keyword in the content that how many times it has been used in that particular webpage. We have extracted this content from the result data with using JSON as mentioned earlier. Appropriate URLs are selected and then classes and sub-classes are defined for the representation of the hierarchy of the desired keyword and its association accordingly. Then it is programs function to determine the relevant words with the main keyword, which means that it does not contain prepositions, etc. and their size does not exceed two characters as well as they are represented in standard ASCII.

6. Each resultant word (candidate) selected is passed through an analysis that includes checking the number count of its occurrence in the web content. After performing analysis, total number count is noted and on that basis an appropriate candidate key is selected.

7. When we have got the resultant word or candidate word, a new keyword is formed by joining two words; candidate word and the main keyword, Facebook Social Network, for instance. Similarly the whole process can be repeated, each recursion can have its own selected candidates, but keeping in view the above mentioned constraint. The recursive process continues till only when there are no results left to be found for the word.

8. We have got the final output in graphical representation of hierarchy of class and sub-class and that is our ontology.

Evaluation and Result

The initial keyword "Network" is selected here for demonstration, and it has been carried along with the constraints mentioned earlier, that minimum size of the chosen word shall be more than two characters, and occurrences (number count) on the web content are two of the few constraints. It can be seen in figure 7, which is visualization of the class hierarchy is protégé. To further elaborate the example, another word i.e. the candidate word broadcast is taken, so the combination of initial and candidate word is "broadcast network", and accordingly the result is produced from various web sources. To analyze it further, the hierarchy is built with different candidate words from broadcast (mainly its types), such as terrestrial, cable, and satellite. The purpose of this research is building the hierarchy of classes using OWL, as it enables to find equalities or inclusions [22], so what we get in result is Internet class contains social network and similarly social network will present the output that is facebook or twitter. Following figure 7 depicts the visualization of OWL file in its editor Protégé.

Figure 7: Network hierarchy as shown in Protégé

This system also stores the relevant URLs and the class names as well, so that user has the comfort for accessing the more relevant websites for the desired keyword. To make the process clearer figure 8 shows the store URLs along with the classes, like mail is subclass of web as well as social network. Jambalaya plug-in of OWL editor Protégé illustrates the complete ontology is various formats. Figure 9, 10, 11 shows the nested tree map, class tree map, and the individual tree map.

Below given is a small part of the OWL file that has been generated by the current mechanism:

<owl:Class rdf:ID="Facebook">

<rdfs:subClassOf

rdf:resource="#Social_Network"/>

</owl:Class>

<owl:Class rdf:ID="FDDI">

<rdfs:subClassOf rdf:resource="#LAN"/>

</owl:Class>

<LAN

rdf:about="http://computer.howstuffworks.com"/>

<Inter_net

rdf:about="http://computer.howstuffworks.com_"/>

<CableCARD

rdf:about="http://en.wikipedia.org/wiki/CableCARD"/

>

<Computer

rdf:about="http://en.wikipedia.org/wiki/Computer_network_"/>

Figure 8: URLs associated with Facebook calss shown in protégé

Figure 9: "Nested Tree Map" as shown under Jambalaya plugin in protégé

Figure 10: "Class and Individual Tree map" as shown under Jambalaya in protégé

Figure 11: "Class Tree" as shown under Jambalaya in protégé

Conclusion

Ontology construction is a growing trend and need of the semantic web as well, and a large number of researchers are working over it in different domains, to name a few structure information working domains, databases, dictionaries, etc, [24, 25], and simultaneously others are getting involved in natural language texts (NLT) processes [26]. Databases play a vital role in construction of ontologies, because they are the primary structured information sources. Taking unstructured data from the web and formalizing it so that it can be structured automatically is a difficult work to do, but apart from its significance it is interesting as well. Through this research it is intended to make the automation public, as it is based on open standard and constructed using publicly available resources of Google, like Google AJAX API and JavaScript parser JSON. The analysis done is here requires more efforts and it can be enhanced through designing algorithms and building complicated relationship of initial and candidate words.

References