❓#DidYouKnow how computers understand the same words with different meanings within a context? something that comes relatively easy to us humans.
💡Word sense disambiguation(#WSD), is a term popularly used in natural language processing, determines the sense or the meaning of a word within a particular context. As humans, it may be an intuitive or unconscious process however, within machines it requires experimentation with various techniques to achieve the understanding of the same words with different meanings.
🤖Although WSD is gaining a lot of attention recently with the fast adoption of #AI techniques, the term was first introduced by Waren Weaver in 1949. The acronym is newer but the investigation of WSD in computational linguistics as a specific task has been going on for more than 40 years now. 🤯
🧐 For example, the ambiguous word "bass" refers to different classes of objects in the following sentence
🗣i) I can hear the bass sound
🗣ii) They like grilled bass
In word sense disambiguation(WSD) this is addressed by mapping a target expression to one of its possible senses with the help of the words surrounding it. The models should map the meaning of low-frequency tones 🎶 and fish 🐟 respectively for the sentences mentioned above.
There are several approaches ⬅️➡️ to WSD and can be broadly classified into the following four:
i) Knowledge-based 📚
ii) Supervised learning 📊
iii) Deep Neural Network 👾
iv) Unsupervised 📰
It also has many applications in various computational linguistics and natural language processing field such as Machine translation 🇦🈂️, Information retrieval ℹ️🔎, various search engines 🔍, Text mining 🗂 and Information extraction📜.
📝 Contributing Editor: Women in AI & Robotics community member Rashmi Kamath