Uses numbers to represent words. For a neural network like word2vec, it uses 300-500 numbers. Once embeddings have been trained, we use them to derive similarities between phrases.
We use mathematical models for processing natural language. We are particularly interested in algorithms that scale well and can be run efficiently in a highly distributed environment.
Probabilistic Context-Free Grammars parsers use knowledge of language gained from hand-parsed sentences to produce the most likely analysis of new sentences.
To approximate NLP understanding, we use a neural network with enough hidden layers.
We use parallel processing in which many calculations are carried out simultaneously to give you the suggestions fast.
We extract plain text embedded in your PDF resumes and cleanse it before parsing.
To support the engineering student community, to provide a test bench for our tools (you would not believe the strange text some people submit), and to promote BYOR.
BYOR asks users for feedback (Do you remember, red, yellow, green butttons on the result page?) on the suggestions it returns. You are participating in making our algorithm better.
We are matching qualified candidates with top tech companies only when candidates explictly give us the permission to do so. We do not sell your resume to recruiters or any other company. We do not associate with any recruiting firm. Unless you give us the permission to match you with a job, your resume and contact information won't be shared with anybody.
No. Resumes uploaded on BYOR cannot be indexed by search engines.
No. We do not offer access to our resume database to anybody.