InformationRetrieval System Evaluation
Informationretrieval can be defined as an activity of acquiring informationresources that are relevant to a particular information need from agroup of information resources. The searches can be on the basis ofeither full-text or metadata indexing. The automated informationretrieval systems play a vital role in reducing information overload.IR systems have been adopted by many public and university librariesin providing access to journals, books, and other documents. The websearch engines such as Google are the most evident IR applications.Information retrieval can be termed as the foundation for the modernsearch engines.
Informationretrieval has become an activity of great importance especially inthe development of contemporary information and communicationtechnologies (Goker & Davies, 2009). Information retrievalevaluation is a very crucial aspect since one gains knowledge onwhich system will be convenient to him or her. Google has a number ofproducts that narrows down the server’s search to become moreaccurate. Google books, is one of the features. It allows servers toaccess books and their reviews easily. Google maps, another productof Google helps people in terms of locating places. The documentbelow evaluates Google as an information retrieval system.
Evaluationcan be defined as ascertaining the amount or value of a certain thingor appraising it. Information retrieval is also a subject toevaluation. The evaluation process involves identification ofsuitable criteria that can be subjected to measurement in one way orthe other. The evaluation process is crucial for the design,development and maintenance of effective information retrievalsystems (Meadow et al., 2007). This is because it enables aninformation retrieval system to be measured and quantified. Accordingto Kowalski & Maybury (2000), the criteria should incorporate theevaluation of the information retrieval system’s characteristics.This includes the system’s retrieval effectiveness or theassessment of the consumers’ satisfaction with the system.
TheInformation retrieval system under evaluation in this document isGoogle. Google is an American-based multinational corporation. Googletends to specialize in internet related products and services whichinclude online advertising technologies, cloud computing, software,and search (Meadow et al., 2007). The founders of Google are SegeyBrin and Larry Page. It was formulated as a private company in theyear 1998. The initial public offering took place later in the year2004. The mission statement of Google was the organization of globalinformation thus making it universally accessible and helpful. Thecorporation is believed to be running more than a million serversglobally in data centers. They also process more than one billionsearch requests as well as approximately twenty-four petabytesuser-generated data every day. Alexa in December 2013 listed Google’swebsite as most visited website in the globe.
Myevaluation on Google as an IR system will entail the criticalanalysis criterion. I used the criterion to evaluate Google on threecontexts. The first aspect is on the basis of its coverage while thesecond one is the recall. Recall entails the ability of Google toretrieve relevant items. The third and the last aspect will beprecision which entails the ability of Google to avoid retrievingirrelevant items. In the critical analysis criterion, I made a searchusing Google as my IR system of choice and analyzed the results onthe three contexts basis. I searched Bahamas from Google.
Onthe coverage aspect, Google gave an impressive coverage of Bahamas.The first eight pages were all analyzing Bahamas. Every page,however, had a unique coverage related to Bahamas. One of the pageswas talking purely on what Bahamas is and how it is has come be theway it is. The other page was analyzing Bahamas as an island showingwhere it is located while the other analyzed Bahamas tourism. Thecontext revealed that Google is 100% effective on the information tobe retrieved since it has a very wide coverage. On the recall aspect,Google’s response is praiseworthy.
Theinformation concerning Bahamas in Google was very substantial.Opening the pages gave me an analysis concerning Bahamas from adifferent perspective. This furnished me with essential informationrelated to Bahamas. All the pages possessed relevant informationconcerning Bahamas. On the precision aspect, Google tried very muchto avoid irrelevant items related to my search. Although Bahamas wasa general search, only one page out of eight pages had informationnot related to Bahamas state. This shows that Google’s precision ispraise worth since the precision is equivalent to 88%. The criterionhelped to get relevant information concerning Google’seffectiveness. Google is an effective information retrieval systemsince the information retrievable is substantial.
Itis worth noting that Google offers the best as far as informationretrieval is concerned. According to the evaluation, Google is notonly efficient in terms of response but also offers qualityinformation from a variety of texts. In terms of efficiency, I foundGoogle can load pages and articles within very short timeframes.Study indicates that within a day, approximately 24 petabytes of datais processed, and millions of servers served. I found Google’sresponse on low network connections as relatively fair. It offersmany free full texts. The evaluation also indicated that most ofGoogle’s information is relevant and offers its consumers withnumerous articles and documents to search from. This reveals the factwhy Google stands to be the most visited website worldwide (Kowalski& Maybury, 2000).
Interms of retrieval capabilities, Google showed a positive response.Since its establishment, Google has recorded a tremendous popularityamong the available web users. This is as a result of its informationretrieval capabilities. Currently, Google answers approximately 150million searches. The fact that approximately three billion webdocuments are indexed currently reveals that more information can beaccessed. The system`s retrieval capability hence is very strong. Itscapability is also harnessed by a collection of articles anddocuments that are available in the system. This indicates that thesystem has the power to offer a solution to nearly all questions anindividual may have (Kowalski & Maybury, 2000).
Google’sretrieval performance is excellent. Its performance is not only basedon efficiency but also extends to the value of information availablefor retrieval. Google has numerous workers in every corner of theglobe who are working tirelessly in updating the availableinformation in the Google’s website. This ensures that allinformation in Google is up to date. This means that the Google’sconsumers will have access to quality and up to date informationhence a remarkable retrieval performance. Google’s performance israted high. According to the evaluation, the system has the potentialof scaling at least hundreds of pages per second, as well as hundredsof millions of pages in a single run. This clearly shows thatGoogle’s retrieval performance is worth praising (Meadow et al.,2010).
Precisionand accuracy are two very essential parameters in the informationretrieval world. This is the measurement of the deviation from theexpected result. Google’s precision has been improving day afterday. It is worth noting that out of eight pages on Google related toBahamas seven of them tallied with the exact information required.This is equivalent to 88%. The evaluation hence revealed thatGoogle’s accuracy and precision are excellent.
Googleresponse to the search is worth appraisal. Within the click of abutton, the search was through. It took 0.5 seconds to load the firstpage concerning Bahamas. Within one and half seconds, Google hadmanaged to load five pages. The good part of it is that even withpoor network Google also managed to respond although at a slowerrate. The advanced search operators are query symbols or words thatdo special actions in the Google search. These operators help theGoogle searchers to find what they are searching accurately andquickly.
Theadvanced search allows one to search phrases or words especially thatlack substantial meaning in the normal Google search. Google also hasnumerous products and features that perform different tasks henceoffering its customers a comprehensive service. Google search is oneof the products that play pivotal roles in the Google’s informationretrieval sector. It acts as the mother of all the other products andallows the servers to access any form of information they want.
Googleadvanced search is another crucial parameter that has made Google apreferable information retrieval system of many. I used a number ofwords and phrases to test the efficiency of the advanced Googlesearch. These phrases included Makkah location, Makkah and the holymosque, Long Island University, James and the giant peach 1996 (filmfrom Disney) and information technology. In accordance to the searchengine, the results gained, the precision and the response time areas analyzed below. After putting the word Makkah in the searchingbar, I got varying results about the word. Some of the results wereabout the hotel while others addressed the word Makkah also referredto as Mecca. I then decided to search Makkah location. The searchgave me precise and exact results about Makkah and also the Makkah’sdimensions. Images were also provided as well as the Makkah’s map
Theprecision can be rated at 100%. All the results in one or the otherrevolved around my search. What differed is the aspect the differentpages were addressing. However, there was no deviation in accuracy asfar as the search was concerned. The response time was 0.69 seconds.I went to my second search, Makkah and the holy mosque. The resultswere perfectly accurate. No deviations at all thus precision were100%. After making use of Boolean and the results, were almostsimilar. Google also furnished me with links in twitter and YouTuberelated to Makkah. The response time was 0.5 seconds.
Thethird search was Long Island University. After putting it on theGoogle bar, it only took 0.43 seconds to respond. Their results werevery accurate without any deviation, and thus the precision was 100%.The number of retrieved results was approximately 59, 900,000although I did not check them all. The first results werespecifically about the Long Island University. I then searched moviefrom Disney, which is “James and the giant peach 1996.” It took0.34 seconds while the precision was rated as 90%. The results foundapproximately one million results. There were few mixed up results,and hence the precision was not perfect. On pressing the image fromthe web pages brought huge images about James, as well as a giantpeach. The images had excellent clarity. On clicking the video,Google linked me to YOUTUP with various movie scenes. The search wasvery interesting especially on clicking the book on the same account,and I received numerous books talking about my subject.
Thefifth search was information technology. I tested it on both theGoogle advance and the Google scholar. The Google advance seemedinteresting and also more educative. It was possible to narrow mysearch. So, that can help you to get the result in a fast time. Forexample, I could choose the language, last update, reading level, andfile type among others. I put in the advance search bar (InformationTechnology), and I narrow it by language which English and narrow itby file type which Pdf. The results were 80% precision. There wasslightly deviation in a number of pages while the response time was(0.75) seconds. The results however were about 130,000,000 results.
TheGoogle’s retrieval performance is also very robust. The system hasthe power of serving millions of servers simultaneously. It alsomanages to tolerate faulty HTML. Any server can hence retrieveinformation from Google irrespective of the form and strength of thenetwork he or she is using. This indicates that Google hasstandardized its retrieval performance such that any individual canretrieve information irrespective of the network condition. Google’sinformation can be accessed any time of the day. This means that itsperformance is available anytime. Any time any person requiresretrieving certain information he or she can access it conveniently.The search engine is set in a manner that no overload can affect itand hence its performance is continuous and guaranteed. The aboveinformation shows that Google’s retrieval performance is excellentand worth praising.
Acomparison of the various search engine indicates that there is adifference that exists between them. Since there are many changedbeing experienced in the internet, it is difficult to come up with astandard evaluation of the Google retrieval system. Engines of Googlesearch like Alta Vista is stronger than other search engines likeLycos and Excite. A historical assessment indicates that searchengines were not in existence before 1994. Although the number ofevaluations conducted on web search engines are not as many, thesearch engines have a considerable impact (Chu and Rosenthal, 1996).
Accordingto the evaluation conducted, Google has come a long way to achievingthe success at hand. Various parameters, aspects, and techniques haveto be incorporated so as to make it efficient. Constant updating ofthe information has helped it to do well in the information retrievalsector. It is worth to note that irrespective of the importanceattributed to information retrieval systems, evaluation of thesesystems is essential so as to understand the best system to us.
Büttcher,S., Clarke, C. L. A., & Cormack, G. V. (2010). Informationretrieval: Implementing
andevaluating search engines.Cambridge, Mass: MIT Press.
Chu,H., & Rosenthal, M. (1996, October). Search engines for the WorldWide Web: A
comparativestudy and evaluation methodology. In PROCEEDINGSOF THE ANNUAL
MEETING-AMERICANSOCIETY FOR INFORMATION SCIENCE Vol.33, pp. 127-
Goker,A., & Davies, J. (2009). Informationretrieval: Searching in the 21st century.Chichester,
Kowalski,G. J., & Maybury, M. T. (2000). Informationstorage and retrieval systems: Theory and implementation.Boston, MA: Kluwer.
Meadow,C. T., Boyce, B. R., Kraft, D. H., & Barry, C. (2007). Textinformation retrieval systems.Amsterdam: Elsevier/Academic Press.