INTERESTING NEWS ABOUT LINGUISTICS
The linguistics technologies behind contemporary search engines is fairly primary. Let me describe what I mean here. Look for search engines are currently included with precise methodology: they provide fast results but without any terminology analysis; they evaluate huge amounts of written text information on the system and the record of concerns that have been fed in.
It’s a way of implementing a current statistical strategy to machine-learning, but the search requirements keep no regards to any analysis of significance.
We’ve got a unique way of working. We want to evaluate the semantic vicinity of the query to the writing it discovers, based on semantic analysis. Sure — it’s more dangerous and it costs more, so less people go down this “knowledge-hungry” path. The search engines and Ms Research have been doing relevant analysis on this things.
Actually, fundamentality was our significant issue when we went forward with the technique we’re using. From the beginning we took the fairly expensive and resource-intensive decision to develop a global terminology design that forced us to follow a specific series. When you work for a terminology, you cannot keep out any of the stages: you need a complete morphology, format, semantics, lexical semantics, and so on.
Creating the terminology design required us a when, and you can add the information of any vocabulary to it. What we designed was a design that validated its own performance, and we reviewed it in five 'languages' — European, British, In german, France and China.
The linguistics technologies behind contemporary search engines is fairly primary. Let me describe what I mean here. Look for search engines are currently included with precise methodology: they provide fast results but without any terminology analysis; they evaluate huge amounts of written text information on the system and the record of concerns that have been fed in.
It’s a way of implementing a current statistical strategy to machine-learning, but the search requirements keep no regards to any analysis of significance.
We’ve got a unique way of working. We want to evaluate the semantic vicinity of the query to the writing it discovers, based on semantic analysis. Sure — it’s more dangerous and it costs more, so less people go down this “knowledge-hungry” path. The search engines and Ms Research have been doing relevant analysis on this things.
Actually, fundamentality was our significant issue when we went forward with the technique we’re using. From the beginning we took the fairly expensive and resource-intensive decision to develop a global terminology design that forced us to follow a specific series. When you work for a terminology, you cannot keep out any of the stages: you need a complete morphology, format, semantics, lexical semantics, and so on.
Creating the terminology design required us a when, and you can add the information of any vocabulary to it. What we designed was a design that validated its own performance, and we reviewed it in five 'languages' — European, British, In german, France and China.
No comments:
Post a Comment