Automatic new topic identification in search engine transaction logs

Özmutlu, H. Cenk; Çavdur, Fatih; Ozmutlu, Seda

doi:10.1108/10662240610673727

Cited by 17 publications

(32 citation statements)

References 31 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The categories of timeintervals are determined with respect to the length of the difference of the arrival times of two consecutive queries. The categorization of timeinterval and search pattern remains similar to those of Ozmutlu and Cavdur (2005a,b) and Ozmutlu et al (2004aOzmutlu et al ( , 2006.…”

Section: Proposed Methodologysupporting

confidence: 53%

“…The initial steps of the methodology, i.e., evaluation by human expert, separating the data into two sets, identification of search pattern, and time interval of each query in the dataset, can be found in previous articles by Ozmutlu et al (2004aOzmutlu et al ( , 2006, and Ozmutlu and Cavdur (2005a,b). However they are also briefly explained in this article for readers' convenience.…”

Section: Proposed Methodologymentioning

confidence: 99%

“…The logic for the automatic search pattern identification can be summarized in Figure 1 (Ozmutlu et al 2004aOzmutlu and Cavdur 2005a,b). We use seven categories of search patterns in this study which are as follows: Ozmutlu et al (2004aOzmutlu et al ( , 2006 and Ozmutlu and Cavdur (2005a,b). We also provide examples for each search pattern: Reformulation: some of the terms of the second query are also included in the first query but the first query has some other terms that are not included in the second query.…”

Section: Proposed Methodologymentioning

confidence: 99%

“…Interpreted in terms of topic shifts, as in Ozmutlu et al (2004aOzmutlu et al ( , 2006 and Ozmutlu andCavdur (2005a, 2005b), precision (P shift ) is the correctly estimated number of shifts by genetic algorithms among all the shifts marked by genetic algorithms (Equation (5)), and recall (R shift ) is the correctly estimated number of shifts by genetic algorithms among all the shifts marked by the human expert (Equation (6)). On the other hand, interpreted in terms of topic continuations, precision (P contin ) is the correctly estimated number of continuations by genetic algorithms among all the continuations marked by genetic algorithms (Equation (7)) and recall (R contin ) is the correctly estimated number of continuations by genetic algorithms among all the continuations marked by the human expert (Equation (8)).…”

Section: Notationmentioning

confidence: 99%

“…Several nonsemantic methodologies, primarily based on learning algorithms, have been previously applied for new topic identification and session identification, such as the Demspter-Shafer Theory (Ozmutlu and Cavdur 2005a;Ozmutlu, Cavdur, and Ozmutlu 2006), neural networks (Ozmutlu and Cavdur 2005b;Ozmutlu, Cavdur, Ozmutlu, and Spink 2004a), but room for improvement exists. A new nonsemantic methodology for fast and successful automatic topic identification would be a valuable contribution to the field of information science.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Defining Topic Boundaries in Search Engine Transaction Logs Using Genetic Algorithms

Ozmutlu

Cosar

2009

Applied Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

This study proposes to use genetic algorithms for defining the topic boundaries in search of engine transaction logs. Users are interested in multiple topics during a search session, and genetic algorithms are used in this study to determine whether a search engine user has changed topics during a session. Sample data logs from FAST and Excite search engines were analyzed. The findings show that genetic algorithms are fairly successful in identifying topic continuations and shifts in search engine transaction logs.It is a challenge to develop effective information retrieval algorithms and interpret user information-seeking behavior, since people have different and changing information needs, and they utilize different informationsearching strategies to solve their information problems (Gremett 2006). One of the challenges in information-seeking behavior is new topic identification or session identification. New topic identification or session identification is discovering when the user has switched from one topic to another during a single search session (He, Goker, and Harper 2002). In order to find useful patterns in user sessions, it is necessary to group the queries on the transaction logs into clusters. After the query clusters have been identified, the common usage patterns of search engine users can be discovered by statistical tools (Huang, Yao, and An 2006).It is important to identify when the user has changed topics to design clustering algorithms and query recommendation algorithms. In addition, it is a very difficult problem to identify user sessions in common access computer centers, such as libraries and PC labs. When a new user accesses

show abstract

Section: Proposed Methodologysupporting

confidence: 53%

Section: Proposed Methodologymentioning

confidence: 99%

Section: Proposed Methodologymentioning

confidence: 99%

Section: Notationmentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Defining Topic Boundaries in Search Engine Transaction Logs Using Genetic Algorithms

Ozmutlu

Cosar

2009

Applied Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

show abstract

Cross‐validation of neural network applications for automatic new topic identification

Özmutlu¹,

Çavdur²,

Ozmutlu³

2007

J. Am. Soc. Inf. Sci.

Self Cite

View full text Add to dashboard Cite

There are recent studies in the literature on automatic topic-shift identification in Web search engine user sessions; however most of this work applied their topic-shift identification algorithms on data logs from a single search engine.The purpose of this study is to provide the cross-validation of an artificial neural network application to automatically identify topic changes in a web search engine user session by using data logs of different search engines for training and testing the neural network. Sample data logs from the Norwegian search engine FAST (currently owned by Overture) and Excite are used in this study. Findings of this study suggest that it could be possible to identify topic shifts and continuations successfully on a particular search engine user session using neural networks that are trained on a different search engine data log.

show abstract

Classifying web search queries to identify high revenue generating customers

Ortiz-Cordova

Jansen

2012

J Am Soc Inf Sci Tec

View full text Add to dashboard Cite

Traffic from search engines is important for most online businesses, with the majority of visitors to many websites being referred by search engines. Therefore, an understanding of this search engine traffic is critical to the success of these websites. Understanding search engine traffic means understanding the underlying intent of the query terms and the corresponding user behaviors of searchers submitting keywords. In this research, using 712,643 query keywords from a popular Spanish music website relying on contextual advertising as its business model, we use a k-means clustering algorithm to categorize the referral keywords with similar characteristics of onsite customer behavior, including attributes such as clickthrough rate and revenue. We identified 6 clusters of consumer keywords. Clusters range from a large number of users who are low impact to a small number of high impact users. We demonstrate how online businesses can leverage this segmentation clustering approach to provide a more tailored consumer experience. Implications are that businesses can effectively segment customers to develop better business models to increase advertising conversion rates. IntroductionMany websites rely on search engines to drive substantial portions of their traffic. Major search engines such as Google, Bing, and Yahoo! use complex algorithms to determine the relevance of a page (Brin & Page, 1998). Websites that appear on the first page of the search results are likely to get more traffic because most users click on first-page results (Jansen & Spink, 2004). These search engines not only drive new visitors, but research has shown that repeat visitors use search engines as navigational tools (Jansen, Spink, & Pedersen, 2005). With search engines being the primary point of entry to the web for many people, the traffic from search engines is vitally important to websites. For online businesses, a visitor to their website could mean a sale, ad revenue, user registration, or exposure to branding.In the context of web searching, the set of terms for which a user searches is called the query. If a user enters a query and then clicks on a result, these query terms are embedded within the URL that is passed from the search engine to the website. This URL is called the referral URL, and the query terms within the referral URL are called the referral keywords. The webpage pointed to by the link the user clicks is called the landing page. Both the referral URL and referral keywords provide important information to the website owner. Examples of such information include where traffic is coming from (i.e., which search engine, for example), what topics searchers are most interested in, and how a particular landing page is indexed by the search engines. Therefore, it is important to understand and study the search keywords and search phrases that are bringing people from the search engines to the websites (Hackett & Parmanto, 2009). When analyzed appropriately, these referral keywords can provide insightful information about user be...

show abstract

Automatic new topic identification in search engine transaction logs

Cited by 17 publications

References 31 publications

Defining Topic Boundaries in Search Engine Transaction Logs Using Genetic Algorithms

Defining Topic Boundaries in Search Engine Transaction Logs Using Genetic Algorithms

Cross‐validation of neural network applications for automatic new topic identification

Classifying web search queries to identify high revenue generating customers

Contact Info

Product

Resources

About