Analysis and Comparison of Search Engines - Essay Example

Contents Contents 1 1.0.Introduction 1 2.0.Data collection 2 2.1.Timeline 2 3.0.Data analysis 2 3.1.Methods: 2 3.1.1.Number of sponsored links per search engine per search term 3 3.2.Figure 1.0: Number of sponsored links per search engine per search term 4 3.2.1.Number of sponsored links relating to the terms searched 5 3.3.Figure 2.0: Number of sponsored links relating to the terms searched 6 3.3.1.Number of the sponsored links appearing as the actual results 7 3.3.2.Terms generating the most and least numbers of sponsored links 7 3.4.Figure 3: Terms generating the most and least numbers of sponsored links 8 4.0.Discussion 8 5.0.Recommendation 10 6.0.Resources 11 1.0. Introduction In the wake of digitalized information, the world has been relying on web search engines so as to satisfy an information need. However, the extent at which different search engines give relevant information vis-à-vis the searched word or topic has been eliciting mixed reactions. Some scholars argue that the level at which a search engine ranks the relevance of each indexed website to a query affects the relevance of the information searched (Brown & Duguid, 2002). However, contemporary researches investigate how different search engines can integrate Sponsored Links with the information searched. Building from the hypothesis, this report seeks to critically assess how www.google.com, www.bing.com, and www.yahoo.com aggressively integrate Sponsored Links with specified topics searched on them. The results obtained will be statistically compared to ascertain relevancy of the Sponsored Links and the topic across the three search engines. 2.0. Data collection 2.1. Timeline As the theory of information retrieval posits, search engines show different features when a topic is searched on. This will be realised from the data collection. Data for this report involve nine (9) different terms. Each term will be searched one at a time a cross the three search engines and characteristics of Sponsored Links appearing on each term per search engine noted. The features the report has analysed are: number of sponsored links per search engine per search term, number of sponsored links relating to the terms searched, ascertaining whether there is any sponsored link appearing as actual search result and the search term that generated the most and least number of Sponsored Links. 3.0. Data analysis 3.1. Methods: Data analysis encompass the process of copy pasting each of the nine (9) terms on each of the search engine then record different features of the Sponsored Links obtained. 3.1.1. Number of sponsored links per search engine per search term This feature aimed at obtaining data regarding the number of sponsored links available when each word is keyed in every search engine. The figure below can help illustrate the findings. 3.2. Figure 1.0: Number of sponsored links per search engine per search term From the figure above, different results are obtained. However, there are some conspicuous results generated by different search engines. For instance, when the term Kevin Rudd was searched across the three search engines, Google and Bing did not generate any Sponsored Links. Similarly, when the term Royal Baby was searched, Google did not have any Sponsored Links. Yahoo on the other hand showed highest number of Sponsored Links with solar power and how to lay paving. Google showed the highest number of search engine with the term Mini pad. Looking at the data generally, www.bing.com has the highest number of Sponsored Links per search engine per search term. According to Chris (2003), the reason why there are fluctuations in the number of Sponsored Links per search engine per search term is the allocations of the heterogeneous objects (search terms) on the page of the search results. 3.2.1. Number of sponsored links relating to the terms searched In this case, data analysis focused on how Sponsored Links related with the term searched. After copy pasting all the 9 terms separately onto every search engine, their relationship with the term was compared. The figure below explains the data obtained. 3.3. Figure 2.0: Number of sponsored links relating to the terms searched An analysis on the number of sponsored links that related to the terms searched gave different results. Interestingly, the term Kevin Rudd did not have any sponsored link relating to it when the three search engines were engaged. According to Manning et al. (2008) terms such as Kevin Rudd are just ordinary names and therefore do not relate so much with any business activity or brand name(s). Across the three search engines, it can be seen that trade items or trademark such as Windows 8 or Solar power showed the highest numbers of sponsored links relating to them. However, this was fluctuating from one search engine to another. In other words, no search engine consistently gave the highest number of sponsored link relating for all terms. Generally, www.google.com topped with number of Sponsored Links relating to the terms searched with a total of 72 related sponsored links. This was followed by www.bing.com at 58 and finally www.yahoo.com at 56. Another unique aspect of the number of Sponsored Links relating to the terms searched compared with number of sponsored links per search engine per search term regards the presentation of the Sponsored Links. For instance, www.google.com presents their Sponsored Links on top, bottom, left and right of the searched item. This was different with www.bing.com which presents the Sponsored Links on top and bottom of the main page. 3.3.1. Number of the sponsored links appearing as the actual results After searching the nine terms in the three different search engines, different results were obtained. For instance, www.google.com showed only two Sponsored Links appearing as the actual results. These were seen through searching ‘how to lay paving’ and ‘mini pad.’ www.yahoo.com showed the highest number which stood at four. The lowest recorded was www.bing.com with one sponsored link appearing as the actual result. 3.3.2. Terms generating the most and least numbers of sponsored links Each of the terms searched across the three search engines generated different numbers of Sponsored Links. The term that generated least number of Sponsored Links was Kevin Rudd generating a total of two Sponsored Links with the three search engines. This was followed by Royal baby generating a total of six Sponsored Links across the three search engines. On the other hand, solar panel was the term that generated the highest number of sponsored links with cumulative figure of thirty six (36) across the three search engines. This was followed by mini ipad standing at thirty one (31). The third highest term was how to lay paving which had a total of twenty-eight (28) Sponsored Links from the three search engines. The figure below can represents how other terms generated sponsored links. 3.4. Figure 3: Terms generating the most and least numbers of sponsored links 4.0. Discussion Results of the data above can be discussed using different approaches and theoretical frameworks. The basis of understanding the relationship between search engines and Sponsored Links calls for the definition of information retrieval (Fisher & Julien, 2009). Information retrieval is the act of obtaining information resources relevant to the information needed (Korfhage 1997). Searching of the relevant information involves metadata however; such are always connected with other links that may be relevant or irrelevant to the needed information. This is what Manning et al. (2008) terms as Ads or Sponsored Links. How does Sponsored Links appear on a site or search term? First, Case (2007) notes that search engines do not reveal how they rank their pages. Secondly, Sponsored Links are advertisement mainly aimed at promoting certain products or services and in most cases paid for per click (Bing, 2007). Looking at the first data, there were different numbers of sponsored links per search engine per search term. www.google.com showed the highest number of search engine with the term mini pad. The strongest reason for this result is that Yahoo on the other hand, showed highest number of sponsored links with solar power and how to lay paving. Google showed the highest number of search engine with the term Mini ipad. The reason noticed is that www.google.com tries to have sponsored links in relevant search results. That is, the term searched will have direct link with the sponsored links. This is what is termed as allocations of the heterogeneous objects on the page of the search results (Brown & Duguid, 2002; Steve & Giles 2004). Another school of thought argues that each search engine is rated differently by clients. In so doing, many clients will prefer to associate their Ads with search engines frequently visited (Maron, 2008). In recently published article, Everingham et al. (2010) describe that www.google.com tend to have many sponsored links on certain search terms as those home pages have done well in Search Engine Optimization (SEO). Another argument concerns the difference in number of sponsored links relating to the terms searched. Looking at the data above, www.google.com leads in number of sponsored links that are directly related with the terms. The strongest argument in this case is that from observation, www.google.com does not allow Ads to be displayed on the results pages if such Ads are irrelevant. The sponsored links appearing on the results pages can also provide essential information. Chris (2003) concretises this finding adding that, “it is possible to find that certain searches may not lead t o any ads at all when it is deemed to compromise the integrity of the search results” (p. 38). These findings have been mentioned further by Everingham, et al (2010). They discuss about diversification and uncertainty as core components that affect the effectiveness of the search results and sponsored links. They argue that some search engine may post no sponsored links depending of the type of the searched term. Looking at the number of the sponsored links appearing as the actual results and terms generating the most and least numbers of sponsored links, Fisher & Julien (2009) explains this difference basing on the effectiveness of the search engines. It can be noted for instance; from his study, that www.google.com offers the highest items on the menu that are visible compared with www.yahoo.com and www.bing.com. It therefore stands out that the relationship between the term and sponsored link is made broader and even the structure becomes deeper (Hock 2007). Actually, this offers the reason why some words such as Kevin Rudd had least number of sponsored links. According to Everingham, et al. (2010) such words has what they term as, “the uncertainty of the expected relevance of returned documents” which greatly reduces when the term is lowly ranked or diversified. 5.0. Recommendation It can be concluded from the analysis above that the three search engines offer different and unique features. For instance, www.bing.com and www.yahoo.com offers the highest number of sponsored links per search term but has no relevance compared with www.google.com. Google also offers the highest number of sponsored links relating to the searched term. This is to mean, the search engine is recommendable for clients looking for more information regarding their searched term. It can also be concluded that unlike the other two engines, Google is unlikely to offer annoying pop up ads. In as much, Yahoo offers the best service for sponsored links appearing as the actual search. Meaning a client can find yahoo useful if not assure about the term. 6.0. Resources Bing, L. (2007), Exploration of Hyperlinks, Web Data Mining: Contents and Data Usage. New York: Springer. Brown, J.S., & Duguid, P. (2002). The social life of information. Boston: Harvard Business Press. Case, D.O. (2007). Looking for information: A survey of research on information seeking, needs, and behavior (2nd ed.). Amsterdam: Elsevier/Academic Press. Chris, S. (2003). Search Engine Birthdays. Retrieved on 5th September, 2012 from http:///www.serachengines.com/html Everingham, M., Van Gool, L., Williams, C. K. I., Winn, J., Zisserman, A. (2010). The PASCAL Visual Object Classes (VOC) Challenge. International Journal of Computer Vision (Springer) 88 (2): 303–338. Fisher, K., & Julien, H. (2009). Information behavior. In B. Cronin (Ed.), Annual review of information science & technology (Vol. 43, pp. 317–358). Medford, NJ: Information Today. Korfhage, R. R. (1997) Information Storage and Retrieval. USA: Wiley. ISBN 978-0-471-14338-3 Hock, R. (2007). “The Extreme Searcher's guide”. Search engine Technol, 3(2):34-39 Manning, C. D., Raghavan, P., Schütze, H., (2008) Introduction to Information Retrieval. Cambridge University Press. Maron, M. E. (2008) A Historical Note on the Origins of Probabilistic Indexing. Information Processing and Management 44 (2): 971–972. Steve, L. & Giles. T. (2004). "Accessibility of information from the Web". Nature400 (6740): 107–9. Read More