In this paper, we describe the IIITH retrieval system used for the ImageCLEF Wikipedia MM task. The system automatically ranks the most similar images to a given textual query. The system preprocesses the data set in order to remove the non-informative terms. For each query, the system finds a ranked list of its most similar images using the textual information only.
Our system uses a combination of the Vector Space Model (VSM) of Information Retrieval and the Boolean model to determine how relevant a given Document is to a User’s query. In general, the idea behind the VSM is the more times a query term appears in a document relative to the number of times the term appears in all the documents in the collection, the more relevant that document is to the query. It uses the Boolean model to first narrow down the documents that need to be scored based on the use of boolean logic in the Query specification.
The score is calculated based on TF-IDF model[
In the TF-IDF model [
tf (t in d) correlates to the term’s frequency, defined as the number of times term t appears in the currently scored document d. Documents that have more occurrences of a given term receive a higher score. The default computation of tf (t in d) in our system is sq.rt(f requency).
idf (t) stands for Inverse Document Frequency. This value correlates to the inverse of docF req (the number of documents in which the term t appears). This means rarer terms give higher contribution to the total score. The default computation of idf (t) in our system is 1+log(numDocs/(docF req+ 1)) 2.2
In a Vector Space Model[
The score of query q for document d correlates to the cosine-distance or dot-product between document and query vectors. A document whose vector is closer to the query vector in that model is scored higher. The scoring we used was the lucene[5] scoring. 3
By using the image’s filename and the text associated with the image, we submitted one run based on the models described in the previous section.
The results for the wikipediaMM task have been computed with the trec eval tool (version 8.1). The submitted runs have been corrected (where necessary) so as to correpond to valid runs in the correct TREC format. The following corrections have been made: • The runs comply with the TREC fomat as specified in the submission guidelines for the task • When a topic contains an image example that is part of the wikipediaMM collection, this image is removed from the retrieval results, i.e., we are seeking relevant images that the users are not familiar with (as they are with the images they provided as examples). • When an image is retrieved more than once for a given topic, only its highest ranking for that topic is kept and the rest are removed (and the ranks in the retrieval results are appropriately fixed). • Ensure that each of the submitted runs has a unique name.
The interpolated recall precision averages are shown in Figure 1, and the summary statistics for the runs sorted by MAP are shown in Table 1.
Run
TXT
NOFB
In this working note, we presented a retrieval system which automatically ranks the most similar images to a given textual query. The system preprocesses the data set in order to remove the noninformative terms. For each query, the system finds a ranked list of its most similar images using the textual information only. Even though our system did not perform well, we are happy that we participated in the ImageCLEF WikipediaMM task and are looking forward to ImageCLEF2009. 5