Mining search logs for usage patterns
Russell-Rose, Tony and Clough, Pauk. 2015. Mining search logs for usage patterns. In: Markus Hofmann and Andrew Chisholm, eds. Text Mining and Visualization: Case Studies using Open-Source Tools. 40 Boca Raton, Florida: CRC Press, pp. 153-172. ISBN 9781482237573 [Book Section]
|
Text
mining search logs for usage patterns.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (1MB) | Preview |
Abstract or Description
One of the greatest opportunities and challenges of the 21st century is the ever increasing significance of data. Data underpins our businesses and our economy, providing awareness and insight into in every sphere of life; from politics to the environment, arts and society. The everyday interactions between people and devices can be harnessed to power a new generation of products and services, allowing us to better understand human needs, aspirations and behaviour.
Of all the data to which we have access, there is none more valuable than the trace people leave when they search for digital information. In browsing the web, people reveal something about their behaviour and habits, but little about their intent. By contrast, when people search for information, they express in their own words their explicit needs and goals. This data represents a unique resource that offers extraordinary potential for delivering insights that can drive the next generation of digital services and applications.
Various studies have been undertaken to understand how and why people interact with search engines. Such studies have led to the creation of frameworks that describe distinct patterns of use, ranging from individual queries to entire information seeking episodes. These patterns may focus on information seeking behavior [9], the types of search tasks that users perform [10], their goals and missions [5], their task switching behavior [4], or the tasks, needs and goals that they are trying to address when using search systems [10, 11].
Moreover, the academic community is not alone in showing an interest in mining search logs. Two highly influential commercial organisations, ElasticSearch and LucidWorks, have both recently released independent logfile analysis platforms (Kibana and SiLK respectively [13, 14]). What unites all of these efforts is the belief that finding distinct, repeatable patterns of behaviour can lead to a better understanding of user needs and ultimately a more effective search experience. In this chapter, we explore the use of data mining techniques to find patterns in search logs, focusing on the application of open source tools and publicly available data.
Item Type: |
Book Section |
||||
Departments, Centres and Research Units: |
|||||
Dates: |
|
||||
Item ID: |
27133 |
||||
Date Deposited: |
14 Oct 2019 09:39 |
||||
Last Modified: |
29 Apr 2020 17:19 |
||||
URI: |
View statistics for this item...
Edit Record (login required) |