Lexalytics Finds Meaning in :-) and LOL
September 25, 2010
We received a news release from Lexalytics. “Lexalytics Unveils Sentiment Analysis of Emoticons, Acronyms; First OEM Engine to Examine Short Form Content for Sentiment Analysis” reveals that the vendor processes non-text emoticons such as :-0. We think this is a good use of available message cues in a short text message. According to the news release:
With the use of emoticons, abbreviations, and confusing “social speak” grammar, micro-blog services such as Twitter present a difficult task for natural language processing systems. These improvements come as part of the yearly software license for Salience, Lexalytics’ core text analytics engine.
Police, intelligence agencies, and Madison Avenue types are likely to give the new capabilities a spin. Will non-text characters illuminate terse, sometimes tokenized messages? We will keep our ear to the ground.
Interesting idea. One question: Has Lexalytics made a breakthrough no other vendor can emulate with a look up table? Do we process messages with these types of content payloads? Not so much.
Stephen E Arnold, September 25, 2010