=Paper=
{{Paper
|id=Vol-1748/paper-18
|storemode=property
|title=Analysing the Role of Key Term Inflections in Knowledge Discovery on Twitter
|pdfUrl=https://ceur-ws.org/Vol-1748/paper-18.pdf
|volume=Vol-1748
|dblpUrl=https://dblp.org/rec/conf/kdweb/HurriyetogluWOB16
}}
==Analysing the Role of Key Term Inflections in Knowledge Discovery on Twitter==
https://ceur-ws.org/Vol-1748/paper-18.pdf
Analysing the Role of Key Term Inflections in
Knowledge Discovery on Twitter
Ali Hürriyetoǧlu1,2 , Jurjen Wagemaker3 ,
Nelleke Oostdijk2 , and Antal van den Bosch2
1
Statistics Netherlands,
P.O. Box 4481, 6401 CZ Heerlen, the Netherlands
a.hurriyetoglu@cbs.nl
2
Centre for Language Studies, Radboud University,
P.O. Box 9103, NL-6500 HD, Nijmegen, the Netherlands
{a.hurriyetoglu,n.oostdijk,a.vandenbosch}@let.ru.nl
3
Floodtags, Binckhorstlaan 36, M2.11, 2511 BE Den Haag, the Netherlands
wagemaker@floodtags.com
Abstract. We introduce our methodology for collecting tweets and iden-
tifying event-related actionable information using key terms and inflec-
tions. The vast amount of user-generated content makes it challenging
to detect relevant information. Therefore, we aim to facilitate extract-
ing morphological, syntactic, and semantic features of a key term semi-
automatically. The results of our study show that handling the inflections
of key terms separately has its advantages: in the disaster scenario that
we are investigating we are successful in discovering relevant and irrele-
vant information effectively.
Keywords: knowledge discovery, inflection analysis, flood, text mining,
machine learning, social media analysis, Twitter
1 Introduction
Twitter provides a platform that contains a rich source of user generated content.
Any user of this platform can post tweets, and language use tends to be rich and
creative. Consequently, increasingly large quantities of novel content is awaiting
proper analysis on Twitter. Especially, in a disaster context, the need of detecting
precise and complete information is in urgent need of a solution.
In this paper, we present our analysis method and apply it on a tweet collec-
tion concerning floods. The method consists of collecting data with a single key
term (a stem, and its inflections), pre-processing, extracting extensive features
for use in machine learning, and determining coherent clusters in tweet subsets.
Human input is used when labeling the clusters.
A central concept in our method is the information thread, i.e, a group of
related tweets. Relatedness is determined by the expert who uses the method.
For example, the word ‘flood’ has multiple senses, including ‘to cover or fill with
2 Hürriyetoǧlu et al.
water’ but also ‘to come in great quantities’4 . A water manager will probably
want to focus on only the water-related sense. At the same time, he will want
to discriminate between different contextualizations of this sense: past, current,
up-coming events, effects, measures taken, etc. By incrementally clustering and
labeling the tweets, the collection is analyzed into different information threads.
Here, we report on the effect that the use of key term inflections has on
knowledge discovery in a tweet collection. In the following sections, we summarize
the related research, the tool we developed to support our methodology, the data
we collected, and the results obtained.
2 Related Studies
Identifying different uses of a word is a word sense induction task [4], which is es-
pecially challenging for tweets [1, 3]. Since a tweet collection potentially contains
innumerable information threads, we propose an incremental-iterative search for
information threads in which we include the human in the loop to determine the
final result. By means of this approach, we can manage the ambiguity of key
terms as well as the complexity of a tweet collection5 .
Detecting relevant information in mass emergencies has considerable com-
plexity and urgency [2]. Therefore, we direct our efforts to contribute to this line
of research without sacrificing precision, speed, and recall.
3 Relevancer
We implemented an open source tool, Relevancer6 . Relevancer supports the anal-
ysis of the tweet collections with the following steps:
Preprocessing Retweets that are identified by the respective JSON field of an
API response and the tweets where ‘RT @’ occurs at the beginning of the
tweet text are eliminated. Moreover, user names and URLs are converted to
‘usrusrusr’ and ‘urlurlurl’ respectively. After that, we exclude exact duplicate
tweets.
Feature extraction Any token that occurs in a tweet text is used as a fea-
ture. Tokens are detected based on a split on a single punctuation mark or
an arbitrary number of space characters. Features can be words, hashtags,
letters, numbers, letter and number combinations, emoticons, or 2, 3, or 4
length punctuation mark combinations.
Near-duplicate detection Tweets that contain the same pentagram (here: 5
consecutive words larger than 2 letters), or have a cosine similarity higher
4
http://www.oed.com/view/Entry/71808
5
The article in the following URL provides an excellent example of the am-
biguity caused by lexical meaning and syntax: http://speld.nl/2016/05/22/
man-rijdt-met-180-kmu-over-a2-van-harkemase-boys/
6
https://bitbucket.org/hurrial/relevancer
Key Term Inflections in Knowledge Discovery 3
than 0.85, based on the features extracted in the previous step, are con-
sidered as a group of near duplicates. We keep only one tweet from each
near-duplicate group of tweets.
Information thread detection Information threads related to the key term
and its inflections are detected using an unsupervised clustering method, viz.
K-Means. Clustering and cluster selection steps are repeated in iterations
until the requested number of coherent clusters is reached. The tweets in the
detected clusters are not included in the following clustering iteration.
Annotation An expert assigns an information thread label to each cluster
based on its coherency and relevancy.
Remaining and new tweets can be clustered or analyzed with the knowledge
discovered in previous iterations of the clustering and annotation process.
4 Use Case: Flood tweet collection
We collected tweets from the public Twitter API7 with the key term ‘flood’
and its inflections ‘floods’, ‘flooded’, and ‘flooding’ between December 17 and 31
2015. We applied the analysis steps supported by Relevancer to each subset.
Detailed statistics of the collected tweets are represented in Table 1. The num-
ber of collected tweets are provided in the column #All. The columns #unique
and unique% contain the counts after eliminating the retweets and duplicate
tweets and the percentage of unique tweets in each subset of the collection.
The unique tweet ratio for the terms ‘flooded’ and ‘flood’ is highest and lowest
respectively.
Table 1. Tweet statistics are presented for the key term used and its inflections
#All #unique unique% clustered relevant irrelevant incoherent
flood 136,295 101,620 75 4,290 3,682 483 125
floods 55,384 47,312 86 2,339 818 1,291 320
flooded 41,545 38,740 94 2,429 1,418 638 283
flooding 77,280 66,920 87 3,003 1,420 1,573 10
Each subset of unique tweets was clustered in order to identify information
threads. The annotation was performed with the labels ‘relevant’, ‘irrelevant, and
‘incoherent’, which are the most general information threads that can be handled
with this approach. We generated 50 clusters for each subset. The number of
tweets that were placed in a cluster is presented in the column clustered.
A cluster is relevant, if it is about a relevant information thread, e.g. an
actionable insight, an observation, witness reaction, event information available
from citizens or authorities that can help people avoid harm. Otherwise, the
7
https://dev.twitter.com/rest/public
4 Hürriyetoǧlu et al.
label is irrelevant. Clusters that are not clearly about any information thread
are labeled as incoherent. Respective columns in Table 1 provide information
about the number of tweets in each thread. Having many incoherent clusters
points towards the ambiguity of a term.
The detailed cluster analysis revealed the characteristics of the information
threads for the key term and its inflections. The key term ‘flood’ (stem form) is
mostly used by the authorities and automatic bots that provide updates about
disasters in the form of a ‘flood alert’ or ‘flood warning’. Moreover, mentioning
the name of the authorities, e.g, flood advisory, enables tweets to fall in the same
cluster. The irrelevant tweets using this term are about ads or product names.
Each inflection of the term ‘flood’ has a different set of uses with a consider-
able number of overlapping uses. The form ‘flooded’ is mostly used for expressing
observations and opinions toward a disaster event and news article related tweets.
General comments and expressions of empathy toward the victims of disasters
are found in tweets that contain the form ‘floods’. Finally, the form ‘flooding’
mostly occurs in tweets that are about the consequences of a disaster.
Another aspect that emerged from the cluster analysis is that common and
specific multi-word expressions containing the key term or one of its inflections,
e.g. ‘flooding back’, ‘flooding timeline’, form at least a cluster around them.
Tweets that contain such expressions can be transferred from the remaining
tweets set to the respective cluster. For example, we identified 806 and 592
tweets that contain ‘flooding back’ and ‘flooding timeline’ respectively.
Finally, named entities, such as the name of a storm, river, bridge, road,
web platform, person, institution, or place, and emoticons enable the clustering
algorithm to detect coherent clusters of tweets containing such named entities.
5 Conclusion and Future Directions
A novel tweet collection and analysis methodology was presented in order to dis-
cover detailed and precise information about disaster events on a microblogging
platform and in a challenging domain. Our results shows that determining and
handling separate uses of a key term and its inflections reveal different angles of
the knowledge we can discover in tweet collections. The majority of the tweets
that were not placed in any cluster can be handled with the knowledge we dis-
covered in the annotation phase, which provides insights about the semantic and
syntactic space each subset covers.
This methodology can be applied to any term in any language. The labeled
tweets can be used to create supervised machine learning models in order to
handle the remaining tweets in the collection as well as new tweets.
Acknowledgments. This research was supported by the Dutch national re-
search programme COMMIT. We gratefully acknowledge the contribution by
Floodtags.
Key Term Inflections in Knowledge Discovery 5
References
1. Gella, S., Cook, P., Baldwin, T.: One Sense per Tweeter... and Other Lexical
Semantic Tales of Twitter. Proceedings of the 14th Conference of the European
Chapter of the Association for Computational Linguistics pp. 215–220 (2014),
http://www.aclweb.org/anthology/E14-4042
2. Imran, M., Castillo, C., Diaz, F., Vieweg, S.: Processing social media messages in
mass emergency: A survey. ACM Computing Surveys (CSUR) 47(4), 67 (2015)
3. Lau, J.H., Cook, P., McCarthy, D., Newman, D., Baldwin, T., Computing, L.: Word
sense induction for novel sense detection. Proceedings of the 13th Conference of the
European Chapter of the Association for computational Linguistics (EACL 2012)
pp. 591–601 (2012)
4. Mccarthy, D., Apidianaki, M., Erk, K.: Word Sense Clustering and Clusterability.
Computational Linguistics 42(2), 4943 (2016)