This ASP.NET project proposes an realistic way to compare the similarity between text fragments, page estimates by using text acquired from web search engines for a couple of words. It explains the various coexistence of words by utilizing the page counts and co-ordinate it with the acquired text fragments from web search engines. To find the similarity between these co-existing words algorithms like novel pattern and pattern clustering algorithm are utilized. The suggested method exceed various thresholds and formerly proposed web based semantic similarity measures on three base data sets showing high correlation with people ratings. Moreover the proposed system significantly improves the accuracy in the community mining task.
AS.NET Project purpose:
Related words of a particular word are listed in manually created portmanteau vocabulary ontologies such as word net. In word net a synset contains a set of synonymous words for a unique set of words .however semantic similarity between entities change over time and across domains.
Project Scope and its Features
There exists a high correlation between word counts obtained from a web search engine. New words are constantly being created and new senses are also assigned to the existing words. Manually maintaining ontologies to acquire these new words and senses is pricey if not unfeasible. The product feature emphasizes on the words dimensional feature vector FPQ.
The author proposes to implement and compute the semantic similarity between words in Search engine without using fragments or Vector Machines. It includes the description of the existing system along with its limitations. It includes the proposed system along with its advantages.
The key considerations for feasibility include economic feasibility, technical feasibility and social feasibility
Hardware and software requirements
The project details the hardware and software requirements.
Project specific requirements
It defines the various requirements such as functional, non-functional and pseudo requirement
This part of the project reviews various books by Resnik and Cilibrasi and Vitanyi and various calculations to calculate the similarity between the two words.
The various modules discussed in the project are
- Lexical pattern extraction.
- Lexical pattern clustering
- Measuring semantic similarity
- Ranking search results
The various data flow diagrams, Use Case diagrams, Activity diagrams, Sequence diagrams, Class diagrams, Flow diagrams and the Project Architecture are discussed
The project explains the various techniques and algorithms used alongwith the screen shots of the important screens
The project uses the .NET framework. The Microsoft .Net framework is explained it detail along with its features. It also lists the languages available to the programmer while using the .net frame work. It describes various classes that can be used in the ASP.net frame work. It explains the objectives of .net framework. The features of SQL-SERVER are also explained in detail. Full code of the project is also explained with the video tutorials.
The code is explained in detail with the help of video clip to install a document
The project describes the various types of testing and the test case tables are also explained in detail. The method proposed outperforms various baselines as well as previously proposed web-based semantic similarity measures, achieving a high correlation with human rating and improved the F-score in a community mining.