Summary of Paper : Taxonomy of websearch by Broder
1 like•438 views
This document summarizes a paper that classified web search queries into three categories: navigational, informational, and transactional. Navigational queries aim to reach a specific website, informational queries seek information on a topic, and transactional queries want to perform an online activity like shopping. The paper found through surveys and query log analysis that around 20-25% of queries were navigational, 40-50% informational, and 25-35% transactional. It also proposed that early search engines only handled informational and navigational queries directly, while third generation engines aimed to better support all query types through semantic analysis and blending external databases.
Summary of Paper : Taxonomy of websearch by Broder
- 1. Review: “A taxonomy of web search” by Audrei Broder, IBM Research Bhavesh Singh 2010CS50281 1. Summary 1.1 Motivation There are three stages in the evolution of web search engines. The very first generation is too close to classic IR (Information Retrieval). A central tenet of classical information retrieval is that the user is driven by an “information need” which is defined as “the perceived need for information that leads to someone using an information retrieval system in the first place”. But the intent behind a web search is not information, it might be navigational (show me the URL of the site I want to reach) or transactional (show me the sites where I can perform a certain transaction, e.g. shop, download a file, or find a map). So the main aim of this paper was to point out this difference and introduce and analyze the classification of web searches. The development of search engines has significant relation with these analysis. Fig. The classic model of IR, augmented for the web. 1|Page
- 2. 1.2 Contribution Their contributions are as follows Classified web queries according to the intent into three classes: 1. Navigational: The immediate intent is to reach a particular site that the user has in mind, either because they visited it in the past or because they assume that such a site exists. 2. Informational: The intent is to acquire some information assumed to be present on one or more web pages in a static form. By static form, they mean that the target document is not created in response to the user query. 3. Transactional: The intent is to perform some web-mediated activity, i.e. the user wants to reach a site where further interaction will happen. This interaction constitutes the transaction defining these queries. The main categories for such queries are shopping, finding various web-mediated services, downloading various type of file etc. In view of the above taxonomy(classification), they identified three stages in the evolution of we search engines: 1. First generation: those uses mostly on-page data (text and formatting) and is very close to classic IR. Supports mostly informational queries. 2. Second generation: those uses off-page, web specific data such as link analysis, anchor-text and click-through data. This generation supports both informational and navigational queries and started in 1998-1999. Google was the first to use link analysis as a primary ranking factor. By now, all major engines use all these types of data. 3. Third generation: This is the emerging generation for the author. It involves attempt to blend data from multiple sources in order to try to answer “the need behind the query”. For instance on a query like ‘San Francisco’ the engine might present direct links to a hotel reservation page for San Francisco, a map server, a weather server, etc. Thus third generation engines go beyond the limitation of a fixed corpus, via semantic analysis, context determination, dynamic data base selection, etc. The aim is to support informational, navigational, and transactional queries. This is a rapidly changing landscape. 1.3 Methodology Author used two methods to determine the prevalence of various types of queries: a survey of Alta Vista users, and an analysis of the query log at Alta Vista. User Survey: The survey window was presented to random users and achieved a response ratio of about 10%. The data discussed there was collected between June 26 and 2|Page
- 3. November 3, 2001 and consisted of 3190 valid returns. The survey questions relevant to this paper and its result were as followsQ2. Which of the following describes best what you are trying to do? 24.53% I want to get to a specific website that I already have in mind 68.41% I want a good site on this topic, but I don’t have a specific site in mind Q3. Which of the following best describes why you conducted this search? 8.16% I am shopping for something to buy on the Internet 5.46% I am shopping for something to buy elsewhere than on the Internet 22.55% I want to download a file (e.g., music, images, programs, etc.) 57.19% None of these reasons Q4. Which of the following describes best what you are looking for? 14.83% A site which is a collection of links to other sites regarding this topic 76.62% The best site regarding this topic Q2 was used to distinguish between navigational and non-navigational queries. The percentage of queries identified as navigational was 24.5%, non-navigational queries accounted for 68.4%, and 7.1% of the surveys did not answer Q1. Thus among respondents to Q2, the percentage of navigational queries was 26.4%. The survey had some additional questions, in particular Q7 was ‘In your own words, please describe the exact piece of information you are seeking.’ Based on sample of 200 queries, the query text, and the explanation provided at Q7, we estimate however that the number of transactional queries among survey respondents is about 36%. Queries that are neither transactional, nor navigational, are assumed to be informational. Log analysis: They selected at random a set of 1000 queries from the daily AltaVista log. From this set we removed non-English queries and sexually oriented queries. (The latter being about 10% of the English queries). From the remaining set the first 400 queries were inspected. Queries that were neither transactional, nor navigational, were assumed to be informational in intent. Type of query Navigational Informational Transactional User Survey 24.5% ?? (estimated 39%) > 22% (estimated 36%) Query Log Analysis 20% 48% 30% 3|Page
- 4. 1.4 Conclusion An understanding of this classification based on ‘the need behind the query’ is essential to the development of successful web search. The search engines of that time (when the paper was written) only deal well with informational and navigational queries, but transactional queries were satisfied only indirectly and hence a third generation in search engines began to emerge. According to author, the main aim of the third generation was to deal efficiently with transactional queries mostly via semantic analyses (understanding what the query is about) and blending of various external databases. 4|Page