A Web App for Identity Search

As mentioned in my earlier post Opening up about sharing the white papers that I had written in past 18 months, here is my second white paper on building an Identity search web app


Abstract

This paper presents a novel approach of doing an identity search. The application shows all the important details in a single window. The details include the person’s educational background, professional experience, his/her organization website and his/her blog. It is a Ruby on Rails based application and is Semantic Web compliant. This paper explains in detail the rationale behind building this application and an in depth details of how this application was built. The paper throws light on the limitations of the current approach and how it can be improved in future so as to make this a ‘serious’ identity search web application.


Introduction

What is common between Facebook, YouTube, Flickr, LinkedIn and Twitter? Well easy to guess, they all are social networking sites which focus on building online community where people share their interest and activities. In Facebook people create their own profile and share their interests through various online communities, YouTube and Flickr are videos and image management and sharing sites, LinkedIn is a business networking site where people create their profile but unlike Facebook the profile contains mainly details about educational background and their profession and use it for serious business networking. Lastly Twitter is a micro-blogging service. In past few years a trend can be observed that majority of data is user generated i.e. information is “by the people, for the people and of the people”. Compared to the era of 1990’s where there were relatively few information producers and a large number of consumers, the situation has changed now and there is lot more focus on an individual, people try to project themselves either through their profiles in Facebook, LinkedIn or by having their personal Weblogs. Thus now there is a personal space for everyone on web.

Let’s see one more aspect of web. It is a vast ocean of information. Currently most of the information on web is represented using natural language (like English, Russian, Hindi, etc) or using graphics multimedia etc. This information can be processed by humans as they can easily form association between disparate forms of data even if they use different terminologies. However the same task is difficult for machines, it is difficult for them to make sense from say an image, draw analogies or to combine information from heterogeneous sources and make associations among them. There are several tasks which involve tedious work of finding, sharing and combining information on the web; it would be nice if such tasks can be automated i.e. performed by machines (intelligent agents).
For ex: Consider an automatic reservation system. The system should

1. know about my preference
2. build up knowledge using past
3. combine the local knowledge with run time services like airline preferences, dietary requirements, calendaring, etc

For making these tasks automated we need a way in which data should be possibly combined, merged on a web scale, there should be some data that describes other data; machines should be able to reason about the data. To make this happen we need a way of defining semantics of the information and services on the web. This concept of making data on web more meaningful is generally referred to as Semantic Web.

Keeping in mind of the above defined two trends of internet i.e. “greater focus on identity of an individual” and “Semantic Web” a Web Application (Web App) has been created. It is based on following premise

1. It is an identity search application. It scans the web and displays the details of an individual like his/her educational background, professional experience, blog, company details etc.

2. It fetches data only from Semantic Web compliant websites. Currently it takes data from websites which uses microformat. The details of microformat will be discussed later in detail.

Related Posts
Part2
Part3