Venkat Java Projects: Mobile:+91 9966499110
Venkat Java Projects: Mobile:+91 9966499110
Sometime some user’s assigns unrelated hash tags to images which make
searching process difficult. To overcome from this issue author has introduce
hash tags filtering technique using which we will filter hash tags to determine
whether hash tag is relevant or irrelevant by matching content of both main
hash tag and the annotator hash tags. If annotator assigns related hash tags
then it will be relevant and supervisor will give good score to that annotator.
Using HIT algorithm we can determine whether that hash tags is used more
frequently or not, if it’s less frequent or unrelated hash tag then we will
consider as stop hash tag.
In paper only one image used so we are also using same image but our
extension technique will extract content from any image but for existing
technique in paper author has expose only image details so we used same
image
Screen shots
Double click on ‘run.bat’ file to get below screen
In above screen I am uploading one image and by seeing that image anybody
can say that cat or kitten sitting on a bed with some stuff and our extension
will describe same sentence or extract same data from image but existing
technique just will check whether given hash tag and annotator tags are
similar or relevant or not relevant. After uploading image will get below
screen
Now click on ‘Run Existing Technique & Get Annotation Rank’ button to get
below screen
In above screen we can see from loaded images above annotations are correct
as image contains cat, doll, cute etc. Existing technique able to extract 8
correct annotation from all annotated text. Now click on ‘Run Extension
Technique & get Automatic Sentence & Annotation’ button to describe
image in sentence and to check extracted words are matching with annotators
words or not.
In above screen in selected text you can see our extension technique
describing image in a sentence and then extracting words from image and
compare with annotator’s tags to get relevant details. Extension technique
able to extract 17 related annotations. Now click on ‘Comparison Graph’
button to get below graph
In above graph x-axis represents technique name and y-axis represents count
of extracted matching annotations and we can see extension technique able to
extract more related words compare to existing technique.
Note: existing technique can able to check with only one image as author
given only one image details in paper and what other images he has used that
information is not available. But extension technique can work with any
image. See another image example
In above image we can see peoples are on beach with umbrellas and
extension technique can extract this information but cannot compare with
existing technique as author not include this image in his annotation dataset
In above screen uploading same image and then click on ‘Run Extension
Technique & get Automatic Sentence & Annotation’ button to get below
results
In above screen in selected text you can see sentence describing image and its
related attributes or hashtag also displaying