We present a system which provides, given a query, a list of semantically related terms. The terms are ranked accordingly to an original semantic similarity measure learned from a huge corpus. The system performs comparably to dictionary-based baselines with no need of any semantic resource such as WordNet. The further study shows that users are completely satisfied with 70% of query results.
Panchenko, A., Fairon, C., Naets, H., & Morozova, O. (2013). Serelex: Search and Visualization of Semantically Related Words. 35th European Conference on Information Retrieval, Moscou, Russie. https://hdl.handle.net/2078.5/159941