Professional Documents
Culture Documents
NLP
NLP
NLP
Natural language processing (NLP) is the ability of a computer program to understand human
language as it is spoken. NLP is a component of artificial intelligence (AI).
Natural Language Processing, or NLP for short, is broadly defined as the automatic
manipulation of natural language, like speech and text, by software.
QUESTION ANSWERING
Question Answering is a computer science discipline within the fields of information retrieval
and natural language processing, which focuses on building systems that automatically
answer questions posed by humans in a natural language. A computer understanding of
natural language consists of the capability of a program system to translate sentences into an
internal representation so that this system generates valid answers to questions asked by an
user [1]. Valid answers mean answers relevant to the questions posed by the user. As the
internal representation of natural language, sentences must adequately map semantics of
this statement, the most natural approach is in the simulation of facts contained in the
sentences using a description of real objects as well as actions and events connected with
these objects. To form an answer it is necessary, in the first place, to execute the syntax and
semantic analysis of a question.
• The TREC-QA dataset contains questions and answer patterns, as well as a pool of
documents returned by participating teams.
• NewsQA dataset is to help the research community build algorithms that are capable of
answering questions requiring human-level comprehension and reasoning skills.
a) IR-based Factoid Question Answering goal is to answer a user’s question by finding short
text segments on the Web or some other collection of documents.
Theorem Proving
Why theorem proving in an AI course?
Reasoning by theorem proving is a weak method, compared to experts systems, because it does not
make use of domain knowledge. This, on the other hand, may be a strength, if no domain heuristics
are available (reasoning from first principles). Theorem proving is usually limited to sound reasoning.
Differentiate between
• a logic (syntax)
• a set of axioms and inference rules
• a strategy on when how to search through the possible applications of the axioms and rules
Examples of axioms
p -> (q->p)
(p->(q->r)) -> ((p->q) ->(p->r))
p \/ ~p
p->(~p->q)
Strategies
forwards - start from axioms, apply rules
backwards - start from the theorem (in general: a set of goals), work backwards to the axioms
depth-first or breadth-first
general questions:
are the rules correct (sound)?
is there a proof for every logical consequence (complete)?
can we remove rules (redundant)?
Having redundant rules may allow shorter proofs, but a larger search space.