PERSONAL WEB REVISITATION BY CONTEXT ANDCONTENT KEYWORDS WITH RELEVANCE FEEDBACK
ABSTRACT
Getting back to previously viewed web pages is a common yet uneasy task for users due to the large volume of personallyaccessed information on the web. This paper leverages human’s natural recall process of using episodic and semantic memory cuesto facilitate recall, and presents a personal web revisitation technique called WebPagePrev through context and content keywords.Underlying techniques for context and content memories’ acquisition, storage, decay, and utilization for page re-finding are discussed. Arelevance feedback mechanism is also involved to tailor to individual’s memory strength and revisitation habits. Our 6-month user studyshows that: (1) Compared with the existing web revisitation tool Memento, History List Searching method, and Search Engine method,the proposed WebPagePrev delivers the best re-finding quality in finding rate (92.10%), average F1-measure (0.4318) and averagerank error (0.3145). (2) Our dynamic management of context and content memories including decay and reinforcement strategy canmimic users’ retrieval and recall mechanism. With relevance feedback, the finding rate of WebPagePrev increases by 9.82%, averageF1-measure increases by 47.09%, and average rank error decreases by 19.44% compared to stable memory management strategy.A time, location, and activity context factors in WebPagePrev, activity is the best recall cue, and context+content based re-findingdelivers the best performance, compared to context based re-finding and content based re-finding.
EXISTING SYSTEM:
To support personal web revisitation, a number of techniquesand tools are developed, including bookmarks,history tools, search engines, metadata annotation andexploitation, and contextual recall systems.Bookmarks. Apart from back=forward buttons, manually/automaticallybookmarking favorite web pages inweb browsers enables users to get back to the previouslyaccessed pages. According to user’s every visited webpage and browsing preferences, built bookmarksautomatically and organized them into a recencylist or layered structure respectively. Gamez etal. further used classifiers to forecast a few of thebookmarks that are more probably to be visited laterand showed them in the browser bookmarks personaltoolbar, so that the user can access the desired webpage through a single mouse click. Bearing similaritiesTo support personal web revisitation, a number of techniquesand tools are developed, including bookmarks,history tools, search engines, metadata annotation andexploitation, and contextual recall systems.Bookmarks. Apart from back=forward buttons, manually/automaticallybookmarking favorite web pages inweb browsers enables users to get back to the previouslyaccessed pages. According to user’s every visited webpage and browsing preferences, built bookmarksautomatically and organized them into a recencylist or layered structure, respectively. Gamez etal. further used classifiers to forecast a few of thebookmarks that are more probably to be visited laterand showed them in the browser bookmarks personaltoolbar, so that the user can access the desired webpage through a single mouse click. Bearing similarities
PROPOSED SYSTEM:
The main contributions of our paper thus lie in thefollowing three aspects:• We present a personal web revisitation technique,called WebPagePrev, that allows users to get back totheir previously focused pages through access contextand page content keywords. Underlying techniques forcontext and content memories’ acquisition, storage, andutilization for web page recall are discussed.• Dynamic tuning strategies to tailor to individual’smemorization strength and recall habits based on relevancefeedback (e.g., weight preference calculation, decayrate adjustment, etc.) are developed for performanceimprovement.• We evaluate the effectiveness of the proposed techniqueWebPagePrev, and report the findings (e.g., the importanceof context and content factors) in web revisitationthrough a 6-month user study with 21 participants.
CONCLUSION
Drawing on the characteristics of human brain memoryin organizing and exploiting episodic events and semanticwords in information recall, this paper presents a personalweb revisitation technique based on context andcontent keywords. Context instances and page contentare respectively organized as probabilistic context treesand probabilistic term lists, which dynamically evolve bydegradation and reinforcement with relevance feedback.Our experimental results demonstrate the effectivenessand applicability of the proposed technique. Our futurework includes 1) prediction of users’ revisitation, 2)extending the technique to support users’ ambiguousre-finding requests, and 3) incorporating social contextfactors in information re-finding.
REFERENCES
[1] A. Cockburn, S. Greenberg, S. Jones, B. Mckenzie, and M. Moyle.Improving web page revisitation: analysis, design and evaluation.IT & Society, 1(3):159–183, 2003.
[2] L. Tauscher and S. Greenberg. How people revisit web pages:empirical findings and implications for the design of historysystems. International Journal of Human Computer Studies, 47(1):97–137, 1997.
[3] J. Teevan, E. Adar, R. Jones, and M. Potts. Information re-retrieval:repeat queries in yahoo’s logs. In SIGIR, pages 151–158, 2007.
[4] M. Mayer. Web history tools and revisitation support: a surveyof existing approaches and directions. Foundations and Trends inHCI, 2(3):173–278, 2009.
[5] L. C. Wiggs, J. Weisberg, and A. Martin. Neural correlates ofsemantic and episodic memory retrieval. Neuropsychologia, pages103–118, 1999.
[6] M. Lamming and M. Flynn. ”forget-me-not”: intimate computingin support of human memory. In FRIEND21 Intl. Symposium onNext Generation Human Interface, 1994.
[7] E. Tulving. What is episodic memory? Current Directions inPsychological Science, 2(3):67–70, 1993.
[8] C. E. Kulkarni, S. Raju, and R. Udupa. Memento: unifying contentand context to aid webpage re-visitation. In UIST, pages 435–436,2010.
[9] J. Hailpern, N. Jitkoff, A. Warr, K. Karahalios, R. Sesek, andN. Shkrob. Youpivot: improving recall with contextual search.In CHI, pages 1521–1530, 2011.
[10] T. Deng, L. Zhao, H. Wang, Q. Liu, and L. Feng. Refinder:a context-based information re-finding system. IEEE TKDE,25(9):2119–2132, 2013.