This paper present the participation of Information Retrieval Lab(IR LAB DA-IICT Gandhinagar) in FIRE 2016 Microblog Track. The main objective of the track is to identify Information Retrieval methodologies to retrieve important information from Twitter posted during the disasters. We have submitted two runs for this track. In the rst run, daiict irlab 1, we have expanded topic term using Word2vec model trained by the tweet corpus provided by the organizer. Relevance score between tweet and corpus are calculated by Okapi BM25 model. Precision@20 ,primary metric, for this run, is 0.3143. In the second run,daiict irlab 2, we have set di erent weight for original term and expanded topic term, we achieve precision@20 around 0.30.
Social media, like Twitter, is a massive source of real-time information. Twitter is, one of the popular micro blogging website, which has massive user-generated content due to its large number of registered user. During the disaster, Twitter proved its importance on many occasions.
In the FIRE 2016 Microblog track [
We started our work by referring TREC MICROBLOG
2015 papers [
CLIP [
University of waterloo [
Given a topic Q v=fF M T1; :::; F M T7g, representing different information needs, given corpus of tweets T=ft1; t2; ::tng, we need to compute the relevance score between tweets and topics.
R score = f (T; Q) 4.
In this section, we discuss the architecture of the proposed system. 4.1
FIRE 2016 Microblog track has given 7 topics. Essentially these topics are our query. We converted these topics into the query by removing stop words and consider Noun proper noun and verb using Stanford POS tagger. 4.2
We have trained Word2vec model [
In this step, non-English tweets were ltered out. Tweet includes smileys, hashtags, and many special characters. We did not consider retweets and tweet with only hashtag or emoticon or special characters. We also ignored the tweet with less than 5 words and removed all the stop words from the tweet. 4.4
In this step, Title and description were merged so as to make topics more informative. To increase the relevance, topics were also pre-processed topics by converting all alphabets to small case and expanding the abbreviations. Example: NYC- New York City. Also, topics were stemmed. Eg: behaving was converted to behave. 4.5
In this phase, we have calculated relevance score between tweets and topics, In the rst run, we kept same weight for original term and expanded term. In the second run, we set weight 2 for the original term in the topics and 1 for expanded term. We have used Okapi BM25 model for calculating relevance score between expanded topics and tweets.
R score = BM 25 Sim(Qexp; T )
We misunderstand the track guideline. We had sent only top 100 tweets for each topic. So our Precision@20 is on the line with other participant but other metric were substantially lower. Table 1 represents the result declared by track organizer. After Getting Gold Standard data from track organizer, we again perform experiments. Table 2 shows the result of top 1000 tweets for each topic.
We have submitted 2 runs in FIRE Microblog track. In the rst run, we expanded topic term by training Word2vec model with corpus provided by track organizer. We have calculated relevance a score between expanded topic term and tweet using Okapi BM25 model. We have kept the same weight of the original term and expanded term. In the second run, we set weight of the original term and expanded the term in the ratio 2 :1. We have put more weight on a word like a^AIJavailablea^AI_, a^AIJrequireda^AI_. After analyzing the results, we conclude that by changing the weight for the original term and the expanded term does not improve Precision@20 but actually have some adverse e ect. However, Recall@1000 improves approximately 2%. 7.