Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Information Retrieval 2012
DOI: 10.1145/2348283.2348330
|View full text |Cite
|
Sign up to set email alerts
|

Language intent models for inferring user browsing behavior

Abstract: Modeling user browsing behavior is an active research area with tangible real-world applications, e.g., organizations can adapt their online presence to their visitors browsing behavior with positive e↵ects in user engagement, and revenue. We concentrate on online news agents, and present a semisupervised method for predicting news articles that a user will visit after reading an initial article. Our method tackles the problem using language intent models trained on historical data which can cope with unseen a… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
7
0

Year Published

2013
2013
2018
2018

Publication Types

Select...
6
1

Relationship

1
6

Authors

Journals

citations
Cited by 8 publications
(7 citation statements)
references
References 25 publications
0
7
0
Order By: Relevance
“…Our experiments mainly prove the point that the referrer domain has a big predictive potential that should be considered when building any browsing user model. As our goal is limited to the prediction of the next page visited, more general content-based techniques for cold-start [1,27] are not directly comparable with our approach, although a more extensive comparison would be valuable to gain a broader view on the problem. At any rate, we hope the findings highlighted in this paper lead to a greater consideration of the referrer domain with particular focus on cold-start problems.…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…Our experiments mainly prove the point that the referrer domain has a big predictive potential that should be considered when building any browsing user model. As our goal is limited to the prediction of the next page visited, more general content-based techniques for cold-start [1,27] are not directly comparable with our approach, although a more extensive comparison would be valuable to gain a broader view on the problem. At any rate, we hope the findings highlighted in this paper lead to a greater consideration of the referrer domain with particular focus on cold-start problems.…”
Section: Resultsmentioning
confidence: 99%
“…There are approaches based on models of similarity among news articles [16,10,20] considering different key factors like textual similarity, recency, coherence, novelty and popularity. Tsagkias and Blanco [26] proposed a language intent model that extracts a query from the user browsing session. In this way they model the user interest using the queries of the users.…”
Section: Related Workmentioning
confidence: 99%
“…Specifically, in the last years there was a surge of interest in the BrowseGraph, a graph where the nodes are web pages and the edges represent the transitions from one page to another made by the navigation of the users. Characterizing the browsing behavior of users is a valuable source of information for a number of different tasks, ranging from understanding how people's search behaviors differ [32], ranking webpages through search trails [1,33] or recommending content items using past history [29]. A comparison between the standard hyperlink graph, based on the structure of the network, with the browse graph built by the users' navigation patterns, has been made by Liu et al [22,23].…”
Section: Browsegraphmentioning
confidence: 99%
“…White et al (2009) examined five types of contextual information in website recommendation while Ye et al (2012) further explored social influence on item recommendation. Moreover, Tsagkias and Blanco (2012) concentrated on analyzing users' browsing behavior on news articles, and Jin (2012) recommended contents through a unified, per-sonalized messaging system. In our work, the accumulated social interaction content is utilized to help determine the interest of future reader.…”
Section: Related Workmentioning
confidence: 99%