Search Engine for unt.edu. (Lab Report Sample)

Overview The implementation of the search engine was done under the proposed model of the vector space retrieval model. In query of the document over the browser, different search engines should get the same results and that is the main aim of the model. The model seeks to ensure that the query run on different search engines bears the same results and tries to map the search results by the user to the exact search terms. The search terms in the search query uses the like option to try and map the very realistic results in the results query. Each document or query is an N-dimensional vector in the Vector Space Model (VSM), where N is the total number of different terms across all documents and queries. The score of the i-th phrase for a vector is contained in the i-th index of that vector. The field of information storage and retrieval (ISR) deals with organizing, organizing, displaying, and recalling bibliographic data. Implementation The implementation of the crawler and indexer were done in JavaScript. A simple UI was designed to allow the users to interact with the tool on the web application and easily maneuver the view the websites. Search term In the process of web crawling, the user should provide an input text to the platform which should crawl and find the specific domains and URLS under the given keywork. A reducer function is applied under this case to counter check and validate the user inputs. export const actionTypes = { SET_SEARCH_TERM : "SET_SEARCH_TERM", }; const reducer = (state , action) => { switch(action.type) { case actionTypes.SET_SEARCH_TERM: return{ ...state , term : action.term }; default : return state; } } export default reducer; The crawling checks all the links in the website link appropriately. All the links on the website gets outsourced using a searching algorithm. The URL search on the website is done step wise by searching all the directories. It checks all the subdirectories on the given directory and lists them. Upon listing the subdirectories, we need to the other directory which now is repeated similarly like the first directory search process. Once all the directories get listed, the analysis of the url extension is done. Some of the subdirectories could be only files like the PDFs and documents and the links gets well-arranged depending on the subdirectory or link. Indexing The indexing is done using the key word with the most character to appear first and in terms of the most searched term on the website or the search engine. The word that has been searched by most people will gain more traffic on the website and hence will get indexed as the first one on the search results. The method is implemented with eh help of the Google APIs method to ensure that the results are real and that the results displayed from the tool are the ones in real time situation. The segment is as below: const useGoogleSearch = (term) => { const [data , setData] = useState (null); useEffect(() => { const fetchData = async() => { fetch( `https://www.googleapis.com/customsearch/v1?key=${API_KEY}&cx=${CONTEXT_KEY}&q=${term}` ).then(response => response.json()).then(result => { setData(result) }) } fetchData(); },[term]) return {data} } Testing Word of choice Unt.edu In performing tests, we used the work unt.edu as the core term which can be accessed and searched by any search engine. The domain is the domain of the institution and was run on different engines the first one being the institutional domain and the other one on the provided system. The tests were conducted on two different cases. The first case is using the implementation web application to run a test using a word of choice as addressed. The word of choice is typed on the search bar and the results checked. In the second case, the test using the word tested in the original implementation was tested in the domain of the institution. The institution domain should give the same results as the ones in the implemented design. After running he two tests, a comparative study was performed to check on the similarity of the two results from different websites. Running on the provided system TO run the tests on the implemented design we first need to execute the design in real time on the website and ensure it is first running. The step of opening the website is first using the node to run the server. We first need to set up the dependencies possible in the system by running the query as below: The figure above shows the execute of the installer script to install all the dependencies. The script installs two packages necessary for the implementation and execution of the scripts well. The ...