Croft, W.B., D. Metzler, T. Strohman. Search Engines: Information Retrieval in Practice. 2015. http://ciir.cs.umass.edu/downloads/SEIRiP.pdf
- Choudhury, Nupur. "World Wide Web and Its Journey from Web 1.0 to Web 4.0." (IJCSIT) International Journal of Computer Science and Information Technologies, Vol. 5 (6) , 2014, 8096-8100. Accessed 13 January 2018, http://ijcsit.com/docs/Volume%205/vol5issue06/ijcsit20140506265.pdf
The field of electronic search is vast. This work helps to explain the basic yet critical aspects of electronic search that continue to have significant impact on digital discovery today. Croft and Metzler also explain the nuances of search engines and how those nuances can affect both information discovery and knowledge acquisition. Access to and accuracy of online artifact discovery risks being the poorer if this kind of knowledge is ignored.
In the context of information retrieval (IR), the authors speak about the basic building blocks of search. They explain that to do a search for relevant texts, engineers need to design algorithms (an algorithm is a set of rules or instructions for solving a problem), “that can compare the text of queries with text in documents…and decide which texts contain the information that is being sought. It is hard to figure out, compare the meaning of a word, or a sentence or a paragraph or a whole story, and bring back relevant results. The authors state that “understanding and modeling how people compare texts, and designing computer algorithms to accurately perform this comparison, is at the core of information retrieval.” In the early days of computing and the Internet, the authors explain how nearly all data was text-based. Today, there are all types of multi-media data types, for example, images, video, audio and graphics. These types of ‘information objects’ have different structures and different kinds of content – making the task of search engines even more difficult to describe and compare among so many different types of information objects. And to make this subject even more confusing, there are so many different types of search :
- Web search (searching the World Wide Web)
- Vertical search (where a web search is performed on a particular content domain
- Enterprise search (searching for information across a huge amount of files located across a corporate internal ‘intra-net’ and in corporate databases.
- Desktop search (searching just the personal desktop of an individual)
- Peer-to-peer search (searching for information in and across networks of computers without any centralized control.
- Ad hoc search (search based on a user query).
Yet no matter what kind of search one is performing, a few key issues always remain fundamental in information retrieval. Once of these is relevance. Another is evaluation, (which is mostly based on precision and recall). The authors define these terms this way: “Precision is a very intuitive measure, and is the proportion of retrieved documents that are relevant. Recall is the proportion of relevant documents that are retrieved.” They define the third core information retrieval issue as “the emphasis on users and their information needs.”
The authors explain that search engines need to support two major functions: the indexing of data and the query process. They say that the ranking component is ‘the core of the search engine’, and define a search engine as “the practical application of information retrieval techniques to large-scale text collections.” Apart from the big issues mentioned above – which all affect the design of search engines, they highlight a number of additional critical features of search engines:
- Performance (such as response time, query throughput, and indexing speed
- Speed of incorporation of new data
- Coverage (meaning degrees of recency or freshness of the ‘age’ of the stored information
- Customizability
- Adaptability
Having explained the rudiments of search and search engines, the chapter focused on Queries and Interfaces has significant implications for my research. Query formulation and re-formulation and results browsing highly determine whether search engines are viewed as effective. They talk about queries representing:
very different information needs and may require different search techniques and ranking algorithms to produce the best rankings.
A query can be a poor representation of the information need. This can happen because the user finds it difficult to express the information need. More often, however, it happens because the user is encouraged to enter short queries, both by the search engine interface and by the fact that long queries often fail.
The authors go into great detail to explain how users can create better queries involving the use of Key Words. This is because many search algorithms are Boolean Logic-based and focus on key words. Towards the end of this chapter, they discuss context and personalization. At their time of writing in 2010, it was accurate for them to have written: “Most contextual information, however, has proved to be difficult to capture and represent in a way that provides consistent effectiveness improvements.” They also explain how to look out for Bias – particularly in click through data.
Today, with semantic search (using machine reading), Google for example, is emphasizing the importance of context. Digital discovery that has migrated to contextual search over keyword search has major implications for the role of artifact discovery on shaping Holocaust Remembrance. Users who understand the fundamentals of information discovery and how search engines work and are changing from keyword to context search, will be more aware of the dangers of search engine bias, and more pro-active in working harder to ask ‘the right question.’
