Information Technology Reference
In-Depth Information
Twitter
Using #jishin (#earthquake) tag to extract activity
sentences which relate to earthquake
e.g. Earthquake M9.0 was just occurred (03-11 14:47)
I am taking refuge at Akihabara (03-11 15:10)
Extract activity attributes
Activity ID (Who, Action, What, Where, When)
act01 (Null, occur, eathquake M9.0, Null
，
03-11 14:47)
act02 (I, take refuge, Null, Akihabara, 03-11 15:10)
act02 becauseOf act01
Convert to RDF/N3
:act01 a :ActionClass ;
:act :occur ;
:what "earthquake M9.0"@en ;
tl:start "2011-03-11T14:47:00"^^xsd:dateTime .
:act02 a :ActionClass ;
:act :take_refuge ;
:where :Akihabara ;
tl:start "2011-03-11T15:10:00"^^xsd:dateTime ;
:becauseOf :act01 .
Fig. 5.
Method of creating semantic data for TiAN
3.2
Creating Semantic Data
Figure 5 explains how to create semantic data for TiAN. We first use #jishin (#earth-
quake) tag which relates to earthquake to collect activity sentences from Twitter.
Secondly, we use our activity extraction method to extract activity attributes, and rela-
tionships between activities. Finally, we convert the extracted data to RDF/N3 to make
semantic data for TiAN (Appendix 1).
4
Prediction of Missing Activity
Let
Can
act
=
is the set of candidate actions of the active user
u
a
at time
t
. Predicting the action of
u
a
at time
t
can be considered as choosing the ac-
tion in
Can
act
, which has the most highest probability. Therefore, we need to calculate
probability of
u
a
did
act
t
at time
t
(
P
u
a
→act
t
).
As shown in Figure 6, we can use collaborative filtering approach (CF) to calculate
P
u
a
→act
t
. However, while traditional CF [15,16,17] is trying to recommend suitable
products on internet for users, our work tries to predict missing action data in real-
world. Different from products, users' actions strongly depend on location, time, and
before-after actions. Additionally, we need to consider execution time of each action.
This means that it is not suitable to use traditional CF for our work.
Below, we propose a novel action-based CF to calculate
P
u
a
→act
t
.
{
act
1
,act
2
, .., act
t
, ...
}
