Language processor

• US 8,660,837 B2
• Filed: 03/15/2010
• Issued: 02/25/2014
• Est. Priority Date: 03/20/2009
• Status: Expired due to Fees

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Funakoshi, Kotaro
• Primary Examiner(s): YEN, ERIC L
• Application Number:
  US12/723,956
• Publication Number:
  US 20100241421A1
• Time in Patent Office: 1,443 Days
• Field of Search: 704/1, 704/9, 704/257
• US Class Current: 704/9
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning

Abstract

A language processor according to the present invention includes a probability calculating section (103) having a probabilistic model by which a probability that a set of partial referring expression pairs represents an object, each pair consisting of an attribute value expression and a part expression modified by the attribute value expression, the set of referring expression pairs including a pair concerning the whole object, is represented by a product of probabilities each of which is a probability that each partial referring expression pair represents the object. The probability that each partial referring expression pair represents the object is obtained using a probability that the object is chosen, a probability that represents the salience of a part of the object, and a probability representing relevance between the part of the object and an attribute value, a probability that the part expression is used for a part of the object, and a probability that the attribute value expression is used for an attribute value. It further includes at least one memory section (101) for storing values of probabilities used for the probabilistic model. The probability calculating section calculates the probability that the set of partial referring expression pairs represents the object using the values of probabilities stored in the at least one memory section and the probabilistic model.

Claims

1. A language processor using a probabilistic model, comprising:

a probability calculating section comprising the probabilistic model, the probability calculating section configured to determine a probability that a set of partial referring expression pairs represents at least one object, wherein the probability is determined by calculating a product of probabilities, wherein each individual probability of the probabilities is a probability that a partial referring expression pair represents one of the at least one object or a part of an object, each pair comprising an attribute value expression and a part expression modified by the attribute value expression,

wherein each individual probability of the probabilities is obtained using a probability that the at least one object is chosen, a salience probability that represents the salience of a part of the at least one object, a probability representing relevance between the part of the at least one object and an attribute value, a probability that the part expression of a given pair is used for a part of the at least one object, and a probability that the attribute value expression of the given pair is used for an attribute value;

at least one memory section configured to store salience probabilities of parts of the at least one object, probabilities representing relevance between the parts of the at least one object and attribute values, probabilities of using part expressions and values of probabilities of using attribute value expressions,

wherein the probability calculating section is configured to calculate the probability that the set of partial referring expression pairs represents the at least one object using values of probabilities stored in the at least one memory section.

2. The language processor according to claim 1, wherein the probability that the at least one object is chosen can be changed according to the situation.

3. The language processor according to claim 1, wherein values of probabilities of using part expressions and values of probabilities of using attribute value expressions are determined in a same manner for an object.

4. The language processor according to claim 1, wherein values of probabilities of using part expressions and values of probabilities of using attribute value expressions are determined individually for different kinds of the at least one object.

5. The language processor according to claim 1, wherein values of probabilities of using part expressions and values of probabilities of using attribute value expressions are determined individually for the at least one object.

6. The language processor according to claim 1, further comprising an object identifying section configured to identify at least one object represented by the set of partial referring expression pairs, the at least one object represented by the set of partial referring expression pairs being among potential objects, wherein probabilities that the potential objects are represented by the set of partial referring expression pairs are obtained using the probability calculating section,

wherein the object identifying section identifies at least one object with a maximum probability among the potential objects as the object represented by the set of partial referring expression pairs.

7. The language processor according to claim 6, further comprising a structure analyzing section configured to generate at least one set of partial referring expression pairs from natural language.

8. The language processor according to claim 7, wherein the structure analyzing section is further configured to generate plural sets of partial referring expression pairs as potential sets of partial referring expression pairs, and wherein the object identifying section is further configured to identify at least one object with a maximum probability that one of the plural sets of partial referring expression pairs represents the object.

9. The language processor according to claim 1, further comprising a referring expression selecting section configured to:

select a set of partial referring expression pairs, among potential sets of partial referring expression pairs, that represents the at least one object; and

send the selected set of partial referring expression pairs to the probability calculating section to obtain a probability that the selected set of partial referring expression pairs represent the at least one object,

wherein the referring expression selecting section is further configured to select a set of partial referring expression pairs having a maximum probability of representing the at least one object, as the set of partial referring expression pairs selected from among the potential sets of partial referring expression pairs.

10. The language processor according to claim 9, further comprising a surface expression realizing section which generates a natural language expression from the set of partial referring expression pairs selected by the referring expression selecting section.

11. A language processor using a probabilistic model, comprising:

a probability calculating section comprising the probabilistic model, the probability calculating section configured to determine a probability

Pr(o|e)

where o is an object, that a set of partial referring expression pairs represents an object,

e={(e₁^p,e₁^a),(e₂^p,e₂^a), . . . , (e_n^p,e_n^a)}

each pair (e_i^p,e_i^a) comprising an attribute value expression and a part expression modified by the attribute value expression, the set of partial referring expression pairs comprising a pair concerning the object, wherein the probability that the object is represented by the set of partial referring expression pairs is determined by calculating a product of probabilities, wherein each individual probability of the probabilities is a probability

Pr(o|e_i^p,e_i^a)

that a partial referring expression pair represents the object, and is represented by a product of a probability the object is chosen, a probability of using the part expression of the pair

Pr(e_i^p|p_u,o)

that the part expression e_i^p is used for a part p_u of the object o, where u is an index for the part p_u, a probability of using the attribute value expression of the pair

Pr(e_i^a|a_v,o)

that the attribute value expression e_i^a is used for an attribute value a_v where v is an index for the attribute value a_v , a salience probability

Pr(p_u|o)

that represents the salience of the part of the object, and a fitness probability

Pr(a_v|p_u,o)

that represents the degree of fitness of the part of the object to the attribute value;

at least one memory section configured to store salience probabilities of parts of the object, fitness probabilities, probabilities of using part expressions and probabilities of using attribute value expressions,

wherein the probability calculating section is configured to calculate the probability that the set of partial referring expression pairs represents the object using values of probabilities stored in the at least one memory section.

12. A probability calculating method for determining a probability that a set of partial referring expression pairs represents an object, each pair comprising an attribute value expression and a part expression modified by the attribute value expression, the set of partial referring expression pairs including a pair concerning a whole of the object, wherein the probability is obtained in a language processor provided with a probability calculating section and at least one memory section, the method comprising the steps of:

for each partial referring expression pair, obtaining, by the probability calculating section, a probability that a partial referring expression pair represents the object, wherein the obtaining of a probability that a partial referring expression pair represents the object comprises using a probability that the object is chosen, a salience probability that represents salience of a part of the object, a probability representing relevance between the part of the object and an attribute value, a probability that the part expression for a given pair is used for a part of the object, and a probability that the attribute value expression for a given pair is used for an attribute value,

wherein values of the probability that the object is chosen, values of the salience probability that represents salience of the part of the object, values of the probability representing relevance between the part of the object and the attribute value, values of the probability that the part expression for the given pair is used for the part of the object, and values of the probability that the attribute value expression for the given pair is used for the attribute value are stored in the at least one memory section; and

obtaining, by the probability calculating section, the probability that the set of partial referring expression pairs represents the object by obtaining a product of probabilities, each individual probability of the product of probabilities is one of the probabilities obtained by the probability calculating section.