DocumentCode
2031773
Title
Towards activity-based app search
Author
Forte, Andrea ; Wei Wang
Author_Institution
AT&T Labs., San Francisco, CA, USA
fYear
2015
fDate
20-22 Jan. 2015
Firstpage
42
Lastpage
47
Abstract
We introduce a novel way to search and recommend mobile applications namely, activity-based app search. Keywords used in activity-based app search are terms such as “drinking coffee”, “during a flight”, “cooking with friends”, “driving”. We define context as What, Where, When and Whom. The results returned by activity-based app search are apps that can be used while performing the activity specified in the search (What) at a certain time (When) in a certain place (Where) within a social context (Whom). The use of any of the returned apps and the activity performed by the user will not interfere with each other. For example, activity-based app search for “cooking” will return apps to use while cooking such as Pandora Internet Radio while a traditional search for “cooking” would return apps for cooking such as cookbook apps. Out of all the many challenges that this new search paradigm introduces, we focus on the activity (i.e., What) as it is the most challenging and novel of all four.
Keywords
Internet; information retrieval; mobile computing; Pandora Internet Radio; activity-based app search; cookbook apps; returned apps; search paradigm; Context; Ear; Internet; Mobile communication; Mobile computing; Mouth; Servers;
fLanguage
English
Publisher
ieee
Conference_Titel
Mobile Computing and Ubiquitous Networking (ICMU), 2015 Eighth International Conference on
Conference_Location
Hakodate
Type
conf
DOI
10.1109/ICMU.2015.7061026
Filename
7061026
Link To Document