=Paper=
{{Paper
|id=None
|storemode=property
|title=Towards a Framework for Adaptive Faceted Search on Twitter
|pdfUrl=https://ceur-ws.org/Vol-823/dah2011_paper_2.pdf
|volume=Vol-823
|dblpUrl=https://dblp.org/rec/conf/ht/AbelCS11
}}
==Towards a Framework for Adaptive Faceted Search on Twitter==
https://ceur-ws.org/Vol-823/dah2011_paper_2.pdf
Towards a Framework for Adaptive Faceted
Search on Twitter
Ilknur Celik1 , Fabian Abel1 , Patrick Siehndel2
1
Web Information Systems, Delft University of Technology
{celik,abel}@tudelft.nl
2
L3S Research Center, Leibniz University Hannover, Germany
siehndel@l3s.de
Abstract. In the last few years, Twitter has become a powerful tool
for publishing and discussing information. Yet, content exploration in
Twitter requires substantial efforts and users often have to scan infor-
mation streams by hand. In this paper, we approach this problem by
means of faceted search. We propose strategies for inferring facets and
facet values on Twitter by enriching the semantics of individual Twit-
ter messages and present different methods, including personalized and
context-adaptive methods, for making faceted search on Twitter more
effective. We conduct a preliminary analysis that shows that semantic
enrichment of tweets is essential for faceted search on Twitter and that
there is essential need for adaptive faceted search on Twitter. Further-
more, we propose an evaluation methodology that allows us to automat-
ically evaluate the quality of adaptive faceted search on Twitter without
requiring expensive user studies.
Key words: faceted search, twitter, semantic enrichment, adaptation
1 Introduction
With the growing information space on the Web and the increasing popularity of
Social Media, Social Web applications became part of daily activities as well as
the source of information for millions of people. The dynamic nature of the Web
and the diversity of the users along with the heavy information load demanded
some form of adaptation or personalization in many Web-based applications
in various domains. Nowadays, many Social Web applications are suffering from
similar information overload problems, where the users of these applications find
it difficult to read, find and follow the relevant and interesting information shared
by a large network of other users. Our research focuses on tackling information
overload in one of the most popular of these applications, Twitter.
Twitter is the most popular micro-blogging site and a growing Social Web
phenomenon that is attracting interest from different types of people all around
the world for a variety of different purposes, such as fast communication, work,
status updates, following news, sports, events, opinions, hot topics, and so on [1–
8]. With millions of Twitter messages (tweets) per day, highly active users are
2 Ilknur Celik, Fabian Abel, Patrick Siehndel
estimated to receive hundreds of tweets every day3 . Due to the lack of any adap-
tive or personalized navigation support in Twitter, users may get lost, become
de-motivated and frustrated in this network of information overload [10]. Ac-
cessing required or interesting fresh content easily is vital in today’s information
age. Hence, there is a need for an effective personalized searching option from the
users’ point of view that would assist them in following the optimal path through
a series of facets to find the information they are looking for, while providing a
structured environment for relevant content exploring. Our research focuses on
investigating ways to enhance searching and browsing in microblogging sites like
Twitter by means of adaptive and personalized faceted search.
Searching and browsing are, indeed, somewhat limited in Twitter. For exam-
ple, one can search for tweets by a keyword or by a user in a timeline that would
return the most recent posts. So, if a user wants to see the different tweets about
a field of sports, and were to search for “sports” in Twitter, only the recent tweets
that contain the word “sports” would be listed to the user. Many tweets that
do not contain the search keyword, but are about different sport events, sport
games and sport news in general, would not be returned. Moreover, the Twitter
keyword search differs from the general Web search due to the restricted message
size of 140 characters in Twitter [9]. Traditional faceted search interfaces allow
users to search for items by specifying queries regarding different dimensions and
properties of the items (facets) [11]. For example, online stores such as eBay4 or
Amazon5 enable narrowing down their users’ search for products by specifying
constraints regarding facets such as the price, the category or the producer of a
product. In contrast, information on Twitter is rather unstructured and short,
which does not explicitly feature facets. This puts constrains on the size and the
number of keywords, as well as facets, that can be used as search parameters
without risking to filter out many relevant results. Hence, searching by more
than one topic (multiple facets), such as “sport events”, would return only those
recent tweets that contain both of these words and miss tweets like “Off to BNP
Paribas at Indian Wells”, which mentions the name and the location of a sport
event without necessarily including the keywords. In this paper, we introduce
an adaptive faceted search framework for Twitter and investigate how to ex-
tract facets from tweets, how to design appropriate faceted search strategies on
Twitter and how to evaluate such a framework. Our main contributions can be
summarized as follows.
Semantic Enrichment We present methods for enriching the semantics of
tweets by extracting facets (entities and topics) from tweets and related
external Web resources.
User and Context Modeling Given the semantically enriched tweets, we pro-
pose user and context modeling strategies that identify (current) interests of
a given Twitter user and allow for contextualizing the demands of this user.
3
http://techcrunch.com/2010/06/08/twitter-190-million-users/
4
http://ebay.com/
5
http://amazon.com/
Towards a Framework for Adaptive Faceted Search on Twitter 3
Adaptive Faceted Search We introduce faceted search strategies for content
exploration on Twitter and propose methods that adapt to the interests and
context of a user.
Evaluation Framework We present an evaluation environment based on sim-
ulated users to evaluate different strategies in our adaptive faceted search
engine on Twitter.
2 Related Work and Our Motivation
The exponential growth of Twitter has attracted significant amount of research
from various perspectives and fields recently. In this section, we focus on the
related work that motivates and inspires our work, as well as relating our work
to the existing literature.
2.1 Content Exploration on Twitter
A prototype for topic-based browsing in Twitter was proposed after observing
how the users manage the incoming flood of updates [10]. This prototype inter-
face, called Eddi, visualizes a user’s Twitter feed using topic clusters constructed
via a topic identification algorithm without using any semantics or natural lan-
guage processing. This approach, however, does not find the relations between
the topics or perform any recommendation of related topics. While it provides
a means for browsing through a user’s own feed by topics, our ambition is to
infer relations between entities of all tweets in the network in order to adapt
the list of facets presented to contain the related entities of the tweet of interest
even outside of the user’s feed. The aim is to provide a means where not only
the users can easily reach to the information they are looking for by controlling
their search parameters as they move along, but can also browse the related
information about the current subject of interest by related people, countries,
cities, events, and other selected facets.
2.2 Semantic Enrichment of Tweets
The main problem in searching microblogging platforms is the size of the mes-
sages. For example, the Twitter messages, with 140 characters limit, are too
short to extract meaningful semantics on their own. Furthermore users tend to
use abbreviations and short-form for words to save space, as well as colloquial
expressions, which make it even harder to infer semantics from tweets. Rowe
et al. mapped tweets to conference talks and exploited metadata of the corre-
sponding research papers to enrich the semantics of tweets to better understand
the semantics of the tweets published in conferences [12]. We follow a similar
approach to this, except we try to enrich the tweets in general and not in a re-
stricted domain like scientific conferences. A study by Kwak et al. revealed that
the majority of the trending topics in Twitter are either headline or persistent
news, with 85% of all the posted tweets being related to news, claiming Twitter
is used more as a news media than a social network [4]. Consequently, we try to
map tweets to news articles on the Web over the same time period in order to
enrich them and to allow for extracting more entities to generate richer facets.
4 Ilknur Celik, Fabian Abel, Patrick Siehndel
2.3 User and Context Modeling for Adaptive Faceted Search in
Twitter
We also try to discover the relations between the extracted entities by studying
different strategies in order to determine relatedness relations between entities
such as persons related to an event and identify any temporal constraints on
such relations. These learnt relations between entities can be utilized to ease
the search by grouping together the related facets and recommending the most
relevant facets that the user is looking for. Marinho et al. proposed a method for
collabulary learning which takes a folksonomy and domain-expert ontology as
input and performs semantic mapping to generate an enriched folksonomy [13].
An algorithm based on frequent itemsets techniques is then applied to learn an
ontology over this enriched folksonomy. A similar approach exploited frequent
itemsets to learn association rules from tagging activities [14]. We study the co-
occurrence frequencies of entity pairs and compare these with other strategies
for tweets in combination with news articles to learn relations between these
entities.
In addition to adapting the facets to the current search, we aim at adapt-
ing the facet values to the current state of the users in order to personalize the
search and content exploration. Liu et al. analyzed content-based recommenders
for Google News and showed that interests in news topics such as technology,
politics, et cetera change over time [15]. They also predicted user interests and
showed that these user profiles in combination with recent trends on Google
News outperform collaborative filtering. Similarly, Chen et al. studied content
recommendation in Twitter and found out that both topic and relevance are im-
portant considerations [16]. They also observed that URLs extracted from the
user’s close social group is more successful than the most popular ones. Corre-
spondingly, we observe the users’ past activities to infer their recent interests
based on their recent tweets and re-tweets. In other words, we build a profile
of user interests in accordance with entities and topics, which is then used to
adapt ranking of the facet values. Re-arranging the facet values according to
user history and interests in line with the trendy topics can accelerate and thus
improve the searching experience.
3 Faceted Search on Twitter
On Twitter, facets describe properties of a Twitter message. For example, per-
sons that are mentioned in a tweet or events a tweet refers to. Oren et al. [11]
formulate the problem of faceted search in RDF terminology. Given an RDF
statement (subject, predicate, object), the faceted search engine interprets (i)
the subject as the actual resource that should be returned by the engine, (ii)
the predicate as the facet type and (iii) the object as the facet value (restriction
value). A faceted query (facet-value pair) that is sent to a faceted search engine
thus consists of a predicate and an object. We follow this problem formulation
proposed by Oren et al. [11] and interpret tweets as the actual resources the
faceted search engine should return. If a tweet (subject) mentions an entity then
Towards a Framework for Adaptive Faceted Search on Twitter 5
!"##$%&'("$#)*' Faceted Search Engine
!"1.//"'4"4)
561"*)738"#)
0$1$#$#' B$%%2,' Semantic Enrichment
+$,"-&,*' facet extraction linkage
./''0$1$#$#'1$-23$#,'4%5&6$#'7%124%'
8$--,'92%':5#'62,':4%,'6;<*==>55///' -.#/,/0))
?/ ''0$1$#$#'@A'B5'82%%2%>'C&4#&D'
849#2%E4F'G$-H$#'02>6&'I4JE''6;///' 12/$3) User and Context Modeling
K/ ''7,'0$1$#$#',L--'14L%>':5#M$#'
561"*)96%$"#)
<#5:$,,25%4-'&$%%2,'<-4)$#'G2#E4'///' !"#$%&'() profile generation relation learning
N/ ''+5>$#'0$1$#$#'3,'O#%4"1'!-$M$%&' *+""*#)
K#1'+5"%1*'82MP-$15%'?Q.Q'R'6&///'' ,-./)*0") !"#$%
S/ 0$1$#$#F'TU5E532J'4%1'+5112JE' 1$--"'*)
#$4J6'&62#1'#5"%1'6;<*==P2&/-)=:///' 2$"-3)
Adaptive Faceted Search
V/ 8$#$')5"',"#<#2,$1'96$%' !"#$%&'%()*%+,+)
0$1$#$#'-5,&'&6$'W/C/'@<$%X*Y$,///'
Z/ 02#,&'M4U5#'&5"#%4M$%&'4[$#'&6$' facet ranking query suggestion
@H'5<$%/'0$1$#$#'4%1'TU5E532J'///'
\/ 865'&62%E,'&64&'+5>$#'0$1$#$#'2,''
(a) Faceted search interface (b) Faceted search architecture
Fig. 1. Adaptive faceted search on Twitter: (a) example interface and (b) architecture
of the faceted search engine.
the type of the entity is considered as facet type (predicate) and the actual
identifier of the entity is considered as facet value (object). For example, given
a tweet t that refers to the tennis player “Federer”, the corresponding URI of
the entity (U RIf ederer ) and the URI of the entity type (U RIperson ) are used to
describe the tweet by means of an RDF statement: (t, U RIperson , U RIf ederer ).
Figure 1(a) illustrates how we envision the corresponding faceted search in-
terface that allows users to formulate faceted queries. Given a list of facet val-
ues which are grouped around facet types such as locations, persons and events,
users can select facet-value pairs such as (U RIevent , U RIwimbeldon ) to refine their
current query ((U RIperson , U RIf ederer ), (U RIsportsgame , U RItennis )). A faceted
query thus may consist of several facet-value pairs. Only those tweets that match
all facet-value constraints will be returned to the user. The ranking of the tweets
that match a faceted query is a research problem of its own and could be solved
by exploiting the popularity of tweets – e.g. measured via the number of re-
tweets or via the popularity of the user who published the tweet (cf. [17]). The
core challenge of the faceted search interface is to support the facet-value selec-
tion as good as possible. Hence, the facet-value pairs that are presented in the
faceted search interface (see left in Figure 1(a)) have to be ranked so that users
can quickly narrow down the search result lists until they find the tweets they
are interested in. Therefore, the facet ranking problem can be defined as follows.
Definition 1 (Facet Ranking Problem). Given the current query Fquery ,
which is a set of facet-value pairs (predicate, object) ∈ Fquery , the hit list H
of resources that match the current query, a set of candidate facet-value pairs
(predicate, object) ∈ F and a user u, who is searching for a resource t via the
faceted search interface, the core challenge of the faceted search engine is to rank
the facet-value pairs F . Those pairs should appear at the top of the ranking that
restrict the hit list H so that u can retrieve t with the least possible effort.
The effort, which u has to invest to narrow down the search result list H,
can be measured by click and scroll operations. Strategies for facet ranking are
discussed in Section 3.2.
6 Ilknur Celik, Fabian Abel, Patrick Siehndel
3.1 Architecture for Adaptive Faceted Search on Twitter
Figure 1(b) illustrates the architecture of the engine that we propose for faceted
search on Twitter. The main components of the engine are the following.
Semantic Enrichment The semantic enrichment layer aims to extract facets
from tweets and generate RDF statements that describe the facet-value pairs
which are associated with a Twitter message. In particular, each tweet is pro-
cessed to identify entities (facet values) that are mentioned in the message. We
therefore make use of the OpenCalais API6 , which allows for the extraction of
39 different types of entities (facet types) including persons, organizations, coun-
tries, cities and events. As Twitter messages are limited to 140 characters, the
extraction of entities from tweets is a non-trivial problem. Thus, we introduced a
set of strategies that link tweets with external Web resources (news articles) and
propagate the semantics extracted from these resources to the related tweets
in [18]. For example, given a tweet “This is great http://bit.ly/2fRds1t”, we
extract entities from the referenced resource (http://bit.ly/2fRds1t) and attach
the extracted entities to the tweet. In our analysis, we show that this semantic
enrichment allows us to significantly better prepare the tweets for faceted search
than enrichment which is merely based on tweets.
User and Context Modeling In order to adapt the facet ranking to the
people who are using the faceted search engine, we propose user modeling and
context modeling strategies. The user modeling strategies model the interests
of the users in certain facet values (entities and topics). We therefore exploit
the tweets that have been published (including re-tweets) by a user. In future
work, we also plan to consider click-through data from the faceted search en-
gine. Context modeling covers mining of new knowledge from the Twitter data.
We therefore propose relation learning strategies that exploit co-occurrence of
entities in Twitter messages to infer typed relationships between entities [19].
Adaptive Faceted Search Based on the semantically enriched tweets, the
learnt relationships between entities extracted from tweets and the user profiles
generated by the user modeling layer, the adaptive faceted search layer solves
the actual facet ranking problem. It provides methods that adapt the facet-
value pair ranking to the given context and user. Furthermore, it provides query
suggestions by exploiting the relations learnt from the Twitter messages. Given
the current facet query, which is a list of facet-value pairs where each value refers
to an entity, we can exploit relationships between entities in order to identify
entities that are related to those entities that occur in the current facet query.
We leave the analysis of such query suggestions for future work. Instead, we
focus on the facet ranking problem and propose different strategies for ranking
facet-value pairs in the next subsection.
6
http://www.opencalais.com/
Towards a Framework for Adaptive Faceted Search on Twitter 7
3.2 Adaptive Faceted Search and Facet Ranking Strategies
Non-Personalized Facet Ranking A lightweight approach is to rank the
facet-value pairs (p, e) ∈ F based on their occurrence frequency in the current
hit list H, the set of tweets that match the current query (cf. Definition 1):
rankf requency ((p, e), H) = |H(p,e) | (1)
|H(p,e) | is the number of (remaining) tweets that contain the facet-value pair
(p, e) that can be applied to further filter the given hit list H. By ranking those
facets that appear in most of the tweets, rankf requency minimizes the risk of
filtering out relevant tweets but might increase the effort a user has to invest to
narrow down search results.
Context-adaptive Facet Ranking The context-adaptive strategy exploits
relationships between entities (facet values) to produce the facet ranking. A
relationship is therefore defined as follows:
Definition 2 (Relationship). Given two entities e1 and e2 , a relationship be-
tween these entities is described via a tuple rel(e1 , e2 , type, tstart , tend , w), where
type labels the relationship, tstart and tend specify the temporal validity of the
relationship and w ∈ [0..1] is a weighting score that allows for specifying the
strength of the relationship.
The higher the weighting score w the stronger the relationship between e1
and e2 . We use co-occurrence frequency as weighting scheme. Hence, given the
enriched tweets, we count the number of tweets both entities (e1 and e2 ) are
associated with. The context-adaptive facet ranking strategy ranks the facet-
value pairs (p, e) ∈ F according to w(ei , e), where ei is a facet value that is
already part of the given query: (pi , ei ) ∈ Fquery (cf. Definition 1):
X
rankrelation ((p, e), Fquery ) = w(ei , e)|(p, ei ) ∈ Fquery (2)
i
Hence, the context-sensitive strategy can only be applied in situations where
the user has already made one selection, so that |Fquery | > 0.
Personalized Facet Ranking The personalized facet ranking strategy adapts
the facet ranking to a given user profile that is generated by the user modeling
layer depicted in Figure 1(b). User profiles conform to the following model and
specify a user’s interest into a specific facet value (entity).
Definition 3 (User Profile). The profile of a user u ∈ U is a set of weighted
entities where with respect to the given user u for an entity e ∈ E its weight
w(u, e) is computed by a certain function w.
P (u) = {(e, w(u, e))|e ∈ E, u ∈ U }
Here, E and U denote the set of entities and users respectively.
8 Ilknur Celik, Fabian Abel, Patrick Siehndel
1
1.2
1x106 1.4
3.6 3.4 3.2 3 2.81.6
1.8 1x106
number of tweets that relate to x facet values
number of tweets that relate to x facet types
2
tweet-based tweet-based
tweet-based + exploitation of news relations tweet-based + exploitation of news relations
100000 100000
10000 10000
1000 1000
100 100
10 10
1 1
1 10 100 1000 1 10
number of facet values (entities) number of facet types
(a) number of facet values per tweet (b) number of facet types per tweet
Fig. 2. Impact of semantic enrichment on (a) the number of facet values per tweet and
(b) the number of distinct facet types per tweet.
Given the set of facet-value pairs (p, e) ∈ F (see Definition 1), the person-
alized facet ranking strategy utilizes the weight w(u, e) in P (u) to rank the
facet-value pairs:
w(u, e) if w(u, e) ∈ P (u)
rankpersonalized ((p, e), P (u)) = (3)
0 otherwise
By combining the above three strategies it is possible to generate further facet
ranking methods. A combination of two strategies can be realized by building the
weighted average computed for a given facet-value pair (p, e) (e.g. rankcombined =
α · rankα ((p, e)) + β · rankβ ((p, e))).
4 Analysis of Faceted Search on Twitter
In our analysis, we study the characteristics of facets on Twitter. As described
above, tweets do not feature many facets by nature. Therefore, strategies that
enrich the semantic of tweets are required in order to derive facet-value pairs
for tweets. In this section, we examine how the semantic enrichment supports
the derivation of facets. Furthermore, we analyze the feasibility of the user and
context modeling strategies for making faceted search on Twitter adaptive.
4.1 Analysis of Semantic Enrichment
As tweets do not provide facets related to the topic, our faceted search frame-
work provides the functionality to enrich the semantics of tweets. To analyze the
feasibility of our semantic enrichment component (see Section 3), we monitored
the Twitter activities of more than 20,000 users over a period of more than two
months and processed the data that we collected (1,671,389 tweets in total) to
extract facet values from the tweets. For 62.91% of the tweets, we succeeded in
extracting at least one entity that we can use as facet value. By making use of the
semantic enrichment functionality that exploits links to external Web resources
(and news articles in particular), we increased the coverage so that 66.77% of
Towards a Framework for Adaptive Faceted Search on Twitter 9
Tweet-only
Tweet+News-based enrichment
10000
distinct entities per user profile
1000
100
10
1
0
1 10 100 1000
user profiles
Fig. 3. Entity-based user profiles that can be exploited for personalized facet ranking.
the tweets which are enriched with facet values obtained from related news have
at least one facet value. In the context of the news-based enrichment, we con-
nected 458,566 Twitter messages with news articles of which 98,189 relations
were explicitly given in the tweets by URLs that pointed to the corresponding
news article. The remaining 360,377 relations were obtained by comparing the
entities that were mentioned in both news articles and tweets as well as com-
paring the timestamps. In previous work we showed that this method correlates
news and tweets with an accuracy of more than 70% [20].
Figure 2(a) reveals that the number of facet values increases clearly when
tweets are enriched with entities of related news articles. For example, less than
20 tweets exhibit more than 10 facet values in the case of semantic enrichment
that is merely based on tweets . Given that tweets are limited to 140 characters,
this observation is expected. Moreover, the number of different facet types per
tweet also increases when linkage to news articles is exploited (see Figure 2(b)).
In our current implementation, we differentiate between 39 different facet types,
where persons, countries and organizations are the most popular types of facets.
In Figure 2(b), we see that the tweet-based enrichment does not allow for more
than 10 different types of facet types per tweets while the exploitation of news
relations features more than 10,000 tweets that can be discovered via more than
10 different facet types, i.e. users can choose between various facets to narrow
down the actual hit list (cf. Figure 1(a)).
4.2 Analysis of User and Context Modeling
The adaptation of the faceted search interface to the preferences of the user and
therefore the personalized facet ranking strategy (see Equation 3) requires entity-
based user profiles (see Definition 3). To analyze to what extent this method can
succeed, we show the profile size of 1500 randomly selected user profiles in Fig-
ure 3. We see that the news-based enrichment results in profiles that provide
more entities than the tweet-only based enrichment. For example, semantic en-
richment based merely on tweets fails for three users as the size of the profile is
zero for these users. In contrast, the news-based enrichment successfully gener-
ates profiles for all users. For more than 98% of the users, the number of distinct
10 Ilknur Celik, Fabian Abel, Patrick Siehndel
entities per profile is even higher than 100. This indicates that news-based en-
richment prevents from sparsity problems and thus allows for supporting the
personalized facet ranking better than the tweets-only-based enrichment.
5 Evaluation Framework for Faceted Search
Evaluating the performance of faceted search is challenging. It usually requires
query logs and click-through data, which is difficult to get for researchers, or
calls for user studies, which are expensive if they are conducted on a large scale.
In this section, we propose a novel technique for automatically evaluating the
performance of faceted search on Twitter. Our evaluation methodology follows
an idea introduced by Koren et al. [21] and exploits re-tweets as ground truth
for estimating user relevance. The evaluation methodology is based on simulated
users who behave in a predefined way. The utility of the interface is measured
by the actions a simulated user needs to perform in order to find a relevant
document.
General Setup. The general setup used for the evaluation process contains
parameters describing the user interface itself and algorithms characterizing the
simulated user behavior. In general, all faceted search user interfaces share some
common characteristics and contains at least two parts: an area displaying the
facets and a part showing the search results. For our evaluation process, the
number of documents to be presented at a time, the number of different facets
to be displayed and the number of elements which can be shown for each different
facet need to be defined. We setup a basic framework for a search interface by
defining these three parameters. Based on this interface, a user can perform
different actions, where the goal is to find a relevant document. For every action
we can define a cost, where the cost is related to the time a real user would
need to accomplish this action. In our scenario a user can perform the following
actions:
Select facet-value pair Basic action a user performs every time a facet-value
pair is clicked, where the displayed search results are automatically updated
after the selection (costs: 1).
View more facet-value pairs This action indicates that none of the currently
displayed facet-value pairs are relevant for the user. By performing this action
the user gets an additional amount of facet-value pairs related to one facet
(costs: 2).
Show more documents This action allows the user to see more documents
(tweets) matching the currently selected facet-values (costs: 2).
Select relevant tweet This action ends the current search (costs: 0).
Beside the actions mentioned above one could also consider the act of dese-
lecting previously marked facet-values. In our search scenario, this action is not
included as we assume that the users have perfect knowledge about the tweet
they are looking for, and therefore a wrong selection will not take place.
Towards a Framework for Adaptive Faceted Search on Twitter 11
Selection Strategies. The simulated users select facet-value pairs based on
different strategies. The strategies we use for our evaluation are:
Random user This user randomly selects one of the displayed facet-values
which matches the tweet he is looking for. If none of the displayed facet-
value pairs matches the tweet, he randomly chooses one facet to see more
facet-value pairs.
First-match user This user selects the first matching facet-value pair dis-
played by the interface. The basic idea behind this strategy is based on
a user who directly clicks on a matching facet-value pair suggestion and do
not look at all displayed facet value pairs to find the best matching one.
Greedy user This strategy tries to reduce the number of matching documents
as fast as possible. This user selects the facet-value pair which occurs in the
least number of remaining documents. This can be motivated by a user who
selects the facet-value pair which is particularly important for the targeted
tweet, in comparison to facet-value pairs which are related to many tweets.
Based on these facet selection strategies, the simulated user searches for a
relevant document. The cost of this search is measured by the costs and number
of actions a user needs to perform to find a relevant document.
Evaluation process. To measure the benefit of the proposed methods for
faceted search, we evaluate the cost for a user to find relevant documents. Here,
a tweet is relevant to a user, if the user re-tweeted this tweet. Re-tweeting a tweet
indicates that the user has read the tweet and is to some extend interested in
the content of the tweet. The proposed method is used to compare the costs of
finding a relevant document when using the baseline ranking strategy based on
frequency (non-personalized facet ranking) in comparison with context-adaptive
facet ranking and personalized facet ranking.
6 Conclusions
In this paper, we presented an adaptive and personalized faceted search engine
for Twitter, where we explained approaches for enriching the semantics of tweets,
extracting facets, discovering relatedness information between entities and ob-
serving user activities to learn their behavior and interests in order to support
users in their search for specific information or tweets. We proposed different
strategies based on learnt relations together with user action history for adapt-
ing the search behavior as well as improving content exploration in Twitter.
Furthermore, we introduced a generic evaluation environment based on Koren
et al. [21] that will allow us to evaluate our strategies by simulated experiments,
which constitutes part of our future research.
Acknowledgements The research leading to these results has received fund-
ing from the European Union Seventh Framework Programme (FP7/2007-2013)
under grant agreement no ICT 257831 (ImREAL project7 ).
7
http://imreal-project.eu
12 Ilknur Celik, Fabian Abel, Patrick Siehndel
References
1. Hughes, A.L., Palen, L.: Twitter Adoption and Use in Mass Convergence and
Emergency Events. In: Proc. of ISCRAM. (2009)
2. Zhao, D., Rosson, M.B.: How and why people Twitter: the role that micro-blogging
plays in informal communication at work. In: Proc. GROUP, ACM (2009) 243–252
3. Cha, M., Haddadi, H., Benevenuto, F., Gummadi, P.K.: Measuring User Influence
in Twitter: The Million Follower Fallacy. In: Proc. of ICWSM, The AAAI Press
(2010)
4. Kwak, H., Lee, C., Park, H., Moon, S.: What is twitter, a social network or a news
media? In: Proc. of WWW, ACM (2010) 591–600
5. Lerman, K., Ghosh, R.: Information contagion: an empirical study of spread of
news on digg and twitter social networks. In: Proc. of ICWSM, The AAAI Press
(2010)
6. Java, A., Song, X., Finin, T., Tseng, B.: Why we twitter: understanding microblog-
ging usage and communities. In: Proc. of WebKDD/SNA-KDD, ACM (2007) 56–65
7. Kaufman, S.J., Chen, J.: Where we Twitter. In: Proc. of Workshop on Microblog-
ging: What and How Can We Learn From It? (2010)
8. Romero, D.M., Meeder, B., Kleinberg, J.: Differences in the mechanics of informa-
tion diffusion across topics: Idioms, political hashtags, and complex contagion on
twitter. In: Proc. of WWW, ACM (2011)
9. Teevan, J., Ramage, D., Morris,M.R.: #TwitterSearch: A Comparison of Microblog
Search and Web Search. In: Proc. of WSDM, ACM (2011)
10. Bernstein, M., Kairam, S., Suh, B., Hong, L., Chi, E.H.: A torrent of tweets:
managing information overload in online social streams. In: Proc. of Workshop on
Microblogging: What and How Can We Learn From It? (2010)
11. Oren, E., Delbru, R., Decker, S.: Extending faceted navigation for rdf data. In:
Proc. of ISWC, Springer (2006) 559–572
12. Rowe, M, Stankovic, M., Laublet, P.: Mapping Tweets to Conference Talks: A
Goldmine for Semantics. In: Proc. of SDoW, colocated with ISWC, CEUR-WS.org
(2010)
13. Balby Marinho, L., Buza, K., Schmidt-Thieme, L.: Folksonomy-based collabulary
learning. In: Proc. of ISWC, Springer (2008) 261–276
14. Hotho, A., Jäschke, R., Schmitz, C., Stumme, G.: Emergent Semantics in BibSon-
omy. In: Informatik für Menschen. Volume 94(2) of LNI, GI (2006)
15. Liu, J., Dolan, P., Pedersen, E.R.: Personalized news recommendation based on
click behavior. In: Proc. of IUI, ACM (2010) 31–40
16. Chen, J., Nairn, R., Nelson, L., Bernstein, M., Chi, E.: Short and tweet: experi-
ments on recommending content from information streams. In: Proc. of CHI, ACM
(2010) 1185–1194
17. Weng, J., Lim, E.P., Jiang, J., He, Q.: Twitterrank: finding topic-sensitive influ-
ential twitterers. In: Proc. of WSDM, ACM (2010) 261–270
18. Abel, F., Gao, Q., Houben, G.J., Tao, K.: Analyzing User Modeling on Twitter
for Personalized News Recommendations. In: Proc. of UMAP, Springer (2011)
19. Celik, I., Abel, F.: Learning Semantic Relationships between Entities in Twitter.
In: Proc. of ICWE, (2011)
20. Abel, F., Gao, Q., Houben, G.J., Tao, K.: Semantic Enrichment of Twitter Posts
for User Profile Construction on the Social Web. In: ESWC, Springer (2011)
21. Koren, J., Zhang, Y., Liu, X.: Personalized interactive faceted search. In: Proc. of
WWW, ACM (2008) 477–486