Word Sense In Graph Essay

Chapter 2

Literature Survey

2.1 General

Word sense disambiguation was one of the important problem during the early days of machine translation. WSD is the task to determine the proper meaning of word and use it in particular context. WSD can be considered as classication problem because the word senses can be the classes.Moreover the automatic classication techniques can be used to recognize and assign each occurrence of the word to classes from external knowledge sources.

2.2 Literature Review

1. Gaona, Gelbukh A Bandyopadhyay advocate to use knowledge based appraches for better word sense disambiguation.

2. El-Rab, Zaiane, El-Hajj stated that use of graph based techniques for WSD proven to be helpful in Biomedical documents.

3. Hessami,

LITERATURE SURVEY 5

4. Che-Yu Yang, Hung stated that use of WordNet library to disambiguate senses by identifying co-occurrence of the words is very helpful.

5. Kolte, Bhirud stated that the sense of the word highly depends on the context n which it is used. This can be easily identied if we know the meaning of other words from the context.

Chapter 3

Techniques for Word Sense

Disambiguation

There are various techniques for word sense disambiguation classied based on source of knowledge used for resolving ambiguity and methods to classify sense of the words.

A. Knowledge Based Techniques

It includes several categories in wsd. This kind of techniques use lexical knowledge source such as dictionary or thesaurus and extract information from word's de nitions and relation with word senses.Few knowledge based techniques are as follows 1) The Lesk Algorithm:[1] in this techniques it nds the overlaps between words that is the number of words that are in common between every denitions of senses. The wordnet dictionary gives following denitions for Ash an Coal.

Ash:-

a) Trees of the olive family with pinnate leaves.

b) The solid residue left when combustible material is throughlt burned or

oxidized.

Also as the raw text can be available in the computable form via WordNet(downloadable/REST

API) , it is also easy to use it in any application. But we can combine Knowledge base with Machine learning approach to improve the accuracy even more.

10

References

1 Gaona, M.A.R.; Gelbukh, A.; Bandyopadhyay, S. "Web-Based Variant of the Lesk

Approach to Word Sense Disambiguation", Eighth Mexican International Conference,

Articial Intelligence, 2009.

2 El-Rab, W.G.; Zaiane, O.R.; El-Hajj, M., "Unsupervised graph-based Word Sense

Disambiguation of biomedical documents", e-Health Networking, Applications &

Services (Healthcom),IEEE 2013

3 Hessami, E.; Mahmoudi, F.; Jadidinejad, A.H.,"Unsupervised weighted graph for

Word Sense Disambiguation", Information and Communication Technologies (WICT),

2011

4 Che-Yu Yang; Hung, J.C.,"Word Sense Determination using WordNet and Sense

Co-occurrence", Advanced Information Networking and Applications, 20th International

Conference, AINA 2006

5 Xiaohua Zhou and Hyoil Han, "Survey of Word Sense Disambiguation Approaches",

American Association for Articial Intelligence