CN109271621B - Semantic disambiguation processing method, device and equipment - Google Patents
Semantic disambiguation processing method, device and equipment Download PDFInfo
- Publication number
- CN109271621B CN109271621B CN201710585495.2A CN201710585495A CN109271621B CN 109271621 B CN109271621 B CN 109271621B CN 201710585495 A CN201710585495 A CN 201710585495A CN 109271621 B CN109271621 B CN 109271621B
- Authority
- CN
- China
- Prior art keywords
- matching
- node
- matching information
- semantic
- text
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 claims abstract description 27
- 238000012545 processing Methods 0.000 claims abstract description 24
- 238000012216 screening Methods 0.000 claims abstract description 17
- 238000001914 filtration Methods 0.000 claims description 18
- 238000004590 computer program Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 13
- 230000008569 process Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 238000003058 natural language processing Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000012925 reference material Substances 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a semantic disambiguation processing method, a semantic disambiguation processing device and equipment thereof, wherein the method comprises the following steps: determining basic matching information of the screened text and generating a plurality of limited matching information associated with the basic matching information according to the corresponding target semantics and ambiguous semantics; establishing a semantic disambiguation data structure for screening target semantic text matching the basic matching information, the data structure comprising: the method comprises the steps of establishing a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node when the matching is successful. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Description
Technical Field
The invention relates to the field of natural language processing, in particular to a semantic disambiguation processing method, a semantic disambiguation processing device and equipment thereof.
Background
Natural language processing is an important direction in the fields of computer science and artificial intelligence. In general, natural language processing, i.e., implementing man-machine natural language communication, or implementing natural language understanding and natural language generation, is very difficult. The underlying cause of the difficulty is the wide variety of ambiguities or ambiguities that exist widely across the various levels of natural language text and dialog.
Specifically, for example, for matching query of keywords to article content, due to the diversity of the language itself, one keyword may have different semantics at the same time, and when text matching is performed using literal content, semantic differences may not be distinguished by the system, so that the matched results may only be consistent literally but not identical semantically.
Disclosure of Invention
The embodiment of the invention provides a semantic disambiguation processing method, a semantic disambiguation processing device and semantic disambiguation processing equipment, and aims to solve the problem that in the prior art, when text matching is performed by using literal content, a system cannot distinguish semantic differences, so that matched results can only be literally consistent but cannot be equivalent to semantically consistent.
The embodiment of the invention provides a semantic disambiguation processing method, which comprises the following steps: determining basic matching information of the screened text, wherein the basic matching information has multiple meanings; generating a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information; establishing a semantic disambiguation data structure for screening matching the basic matching information and conforming to a target semantic text, the data structure comprising: and establishing a relation set of a parent node and child nodes corresponding to the root node and the plurality of child nodes according to the target semantics and the ambiguous semantics, wherein the text to be matched is matched with the matching information corresponding to the current parent node, and if the matching is successful, whether the matching result with the parent node is reversed is determined according to the matching result with the corresponding child node.
Another embodiment of the present invention provides a semantic disambiguation processing apparatus, including: the determination module is used for determining basic matching information of the screened text, wherein the basic matching information has multiple meanings; a generating module, configured to generate multiple pieces of limited matching information associated with the basic matching information according to a target semantic and an ambiguous semantic corresponding to the basic matching information; an establishing module for establishing a semantic disambiguation data structure for screening semantic texts matching the basic matching information and conforming to a target, the data structure comprising: and establishing a relationship set of a parent node and child nodes corresponding to the root node and the child nodes according to the target semantics and the ambiguous semantics, wherein the text to be matched is matched with the matching information corresponding to the current parent node, and if the matching is successful, whether the matching result with the parent node is reversed is determined according to the matching result with the corresponding child node.
Yet another embodiment of the present invention provides a server, including: a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the semantic disambiguation processing method according to the embodiment of the first aspect of the invention when executing the computer program.
A further embodiment of the present invention provides a storage medium, configured to store an application program, where the application program is configured to execute the semantic disambiguation processing method according to the first aspect of the present invention.
The technical scheme provided by the embodiment of the invention has the following beneficial effects:
determining basic matching information with multiple meanings of a screened text, generating a plurality of limited matching information associated with the basic matching information according to target semantics and ambiguity semantics corresponding to the basic matching information, and finally establishing a semantic disambiguation data structure which is used for screening the text matched with the basic matching information and conforms to the target semantics, wherein the data structure comprises: the method comprises the steps of obtaining basic matching information, obtaining a plurality of limiting matching information, establishing a root node corresponding to the basic matching information and a plurality of child nodes corresponding to the limiting matching information, establishing a relation set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and determining whether to reverse a matching result with the parent node or not according to a matching result with the corresponding child node when the matching is successful. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect can be achieved.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
FIG. 1 is a flow diagram of a semantic disambiguation processing method according to one embodiment of the invention;
FIG. 2 is a schematic diagram of a semantic disambiguation data structure according to one embodiment of the invention;
FIG. 3 is a flow diagram of a semantic disambiguation processing method according to another embodiment of the invention;
FIG. 4 is a schematic diagram of a semantic disambiguation data structure according to another embodiment of the invention;
FIG. 5 is a schematic diagram of a semantic disambiguation data structure according to yet another embodiment of the invention;
FIG. 6 is a schematic diagram of a concrete semantic disambiguation data structure according to yet another embodiment of the invention;
FIG. 7 is a schematic structural diagram of a semantic disambiguation processing apparatus according to a first embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a semantic disambiguation processing apparatus according to another embodiment of the present invention;
FIG. 9 is an interaction flow diagram of a semantic disambiguation processing method according to an embodiment of the invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
The following describes a semantic disambiguation processing method, apparatus, and device according to an embodiment of the present invention with reference to the accompanying drawings.
At present, keyword matching is directly identified by completely matching keywords with a sliced part of an article, and most of the attention is focused on how to match a large number of keywords more efficiently, and semantic matching is rarely involved.
In the prior art, one keyword may have different semantics at the same time, and when the literal content is used for text matching, the matching result is only literally consistent but not equal to the semantic consistency due to the indistinguishable semantic differences, i.e., the optimal semantic matching result cannot be obtained.
In order to solve the above problems, in the semantic disambiguation processing method according to the embodiment of the present invention, after the basic matching information with multiple meanings is determined, multiple limited matching information associated with the basic matching information is generated according to the target semantics and the ambiguous semantics corresponding to the basic matching information, and finally, a semantic disambiguation data structure for screening the text matching the basic matching information and conforming to the target semantics is established, that is, the semantics corresponding to the basic matching information not meeting the intention can be filtered out through the semantic disambiguation data structure, so as to achieve a more accurate semantic matching result.
The semantic disambiguation processing method according to the embodiment of the present invention will be described in detail below with reference to the drawings and specific embodiments.
Fig. 1 is a flowchart of a semantic disambiguation processing method according to an embodiment of the present invention, as shown in fig. 1, the semantic disambiguation processing method including:
step 101, determining basic matching information of the screened text, wherein the basic matching information has multiple meanings.
Specifically, in practical applications, there are many forms of basic matching information, and the basic matching information can be selectively set according to practical application needs, for example, as follows:
a first example, a keyword string.
A second example, regular expression.
More specifically, as an example, the form of the basic matching information is determined as a keyword string "golf" in the form of a filter text "golf car related information"; as another example, the screening text is "find reference material for comparing various performances of the mobile phone", the basic matching information is determined in the form of regular expression "know more mobile phone performance reference material", and so on.
It should be noted that the basic matching information has multiple meanings, such as the above example basic matching information "golf", and the semantics thereof may be "playing golf", "golf" and "golf car", etc. In order to match more and accurate results, the basic matching information is "up" to ensure that more contents can be covered for matching.
And 102, generating a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information.
Specifically, the basic matching information has a plurality of semantics, including a target semantic and an ambiguous semantic, such as continuing with the example of the basic matching information "golf" in the above example, whose ambiguous semantics may be "golf" and "golf"; the target semantic may be "golf car".
Further, a plurality of limited matching information associated with the basic matching information may be generated in a variety of forms according to the target semantics and the ambiguous semantics corresponding to the basic matching information, for example, as follows:
in a first example, the qualified matching information is generated according to a preset matching algorithm.
In a second example, qualified match information is generated based on preset stop words.
In a third example, the qualified match information is generated according to a preset context range.
Specifically, one or more of the above manners of generating the limited matching information may be selected according to the actual application requirement. It can be understood that the number of matching information and the diversity of generation modes are limited, and the performance of the semantic disambiguation data structure can be enriched.
Specifically, continuing with the basic matching information "golf" in the above example as an example, the generation of the limited matching information according to the preset matching algorithm may be "golf game grand, golf game download, golf stand-alone game, golf game recommendation", the generation of the limited matching information according to the preset context range as "golf professional site whose golf channel is most authoritative for grand sight sports", and so on.
Step 103, establishing a semantic disambiguation data structure for screening the text matched with the basic matching information and conforming to the target semantic, wherein the data structure comprises: the method comprises the steps of setting a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and if matching is successful, determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node.
Specifically, a semantic disambiguation data structure for screening semantic text matching the basic matching information and conforming to the target semantic is established. The semantic disambiguation data structure has various forms, and can be selectively set according to the actual application needs, for example, as follows:
in a first example, a json structure is used to represent a semantic disambiguation data structure.
The second example, represents the semantic disambiguation data structure using an XML structure.
It is to be understood that the established data structure includes a root node corresponding to the basic match information, a plurality of child nodes corresponding to the plurality of qualified match information, and a set of relationships between parent and child nodes corresponding to the root node and the plurality of child nodes established according to the target semantics and the ambiguous semantics.
In order to make the description of the above structure more clear to those skilled in the art, the description is given in conjunction with fig. 2 as a specific example as follows:
FIG. 2 is a schematic diagram of a semantic disambiguation data structure according to one embodiment of the invention, as shown in FIG. 2:
specifically, the semantic disambiguation data structure comprises: a root node A corresponding to the basic matching information; a plurality of child nodes B, C, D and E corresponding to the plurality of qualified matching information; and establishing a set of relationships between parent nodes and child nodes corresponding to the root node a and the plurality of child nodes B, C, D and E according to the target semantics and the ambiguous semantics.
More specifically, root node a is the parent node with two child nodes B and C; b as a parent node with two child nodes D and E; there are no nodes behind C.
Further, matching the text to be matched with the matching information corresponding to the current parent node, and determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node when the matching is successful, for example, when any one of two child nodes B and C of a is successfully matched with the text, the matching of a is failed, or when any one of two child nodes D and E of B is successfully matched with the text, the matching of B is failed. And in case that B and C fail, the matching of A is considered to be successful. It will be appreciated that the parent node matches successfully only if none of the child nodes match unsuccessfully.
It should be noted that if the matching fails, the non-matching is directly identified.
In summary, the semantic disambiguation processing method according to the embodiment of the present invention determines the multi-semantic basic matching information of the screened text, generates a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguity semantics corresponding to the basic matching information, and finally establishes a semantic disambiguation data structure for screening the text matching the basic matching information and conforming to the target semantics, where the data structure includes: the method comprises the steps of setting a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node when the matching is successful. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Based on the above detailed description of the embodiments, it can be known that there are many semantic disambiguation data structures for screening the semantic text matching the basic matching information and conforming to the target semantic text, and in order to make the semantic disambiguation data structures more clear to those skilled in the art, the following description will be given by taking a json structure as an example to represent the semantic disambiguation data structure:
specifically, a json structure is adopted to represent the semantic disambiguation data structure as follows:
[
{
' word ' golf ' is provided,
‘invert’:[
{
' word ' golf ' is provided,
‘invert’:null
},
{
' word ' golf ' is provided,
‘invert’:null
}
]
},
…
]。
the method comprises two parts, namely basic matching information word and invert representing matching inversion logic. "word" represents basic matching information, i.e., a portion that needs to be matched in text. "invert" refers to a collection of non-keyword sense-compliant logic, where the logic may be a non-word sense compliant keyword or other logic. For example, the content of the car representing "golf" needs to be matched, but the word "golf" is ambiguous, so after the basic matching information "golf" is configured, two words representing ambiguous meanings are additionally configured in the set of "invert", which means that if the word in "invert" is matched, the result that "golf" is matched is regarded as that basic matching information "golf" is not matched.
Specifically, the inverted invert set is also composed of a series of device cells including "word" and "invert". Thus, each inverted logic itself may also be inverted again and so forth recursively. Until the "invert" of a certain device unit is empty, it indicates that the logical link table is bottomed. For example, in the above example, the "golf ball" in the reverse rotation is also a separate device unit, but since its "invert" is empty, it means that there is no additional reverse logic trigger after the "golf ball" is matched, and thus the "golf ball" is considered a valid match.
Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Based on the embodiment, a complete matching semantic disambiguation data structure is loaded and constructed in the system, basic matching information is marked as a root node, a depth-first traversal strategy is adopted for the text to match the whole semantic disambiguation data structure, and a semantic matching result of the keyword is returned. Described in detail with reference to fig. 3 is as follows:
fig. 3 is a flowchart of a semantic disambiguation processing method according to another embodiment of the present invention, as shown in fig. 3, after the above embodiment, the semantic disambiguation processing method further includes:
step 201, obtaining a text to be matched, starting from basic matching information corresponding to a root node of a semantic disambiguation data structure, adopting a depth-first traversal strategy for the text, and matching limited matching information corresponding to a related node in the semantic disambiguation data structure, wherein the depth-first traversal strategy comprises: filtering logic and backtracking logic.
Step 202, determining the matching results of all child nodes of the root node according to the matching results of the related nodes, and outputting semantic matching results corresponding to the text and the basic matching information of the root node.
Specifically, the semantic disambiguation data structure can be regarded as a multi-branch tree, each parent node contains basic matching information attributes and represents the matching logic of the node itself, and the semantic disambiguation data structure further comprises a plurality of child nodes which play a role in reversing the matching result of the parent node.
As an example, as shown in FIG. 4, the root node 1 represents a new basic match information having three child nodes 2, 7, 8 representing inverted semantics, wherein the child node 7 no longer contains any child nodes, thus representing the end of the child node 7 on this branch. The child nodes 2, 8 still include their respective inverted child nodes, so that when there is successfully matched logic in the child nodes 2, 8, it is necessary to go further down to a deeper level for inversion logic determination.
Therefore, as shown in fig. 4, to determine whether a node is finally successfully matched, it is necessary to synthesize the matching results of the node and all child nodes of the node.
It can be understood that, judging whether a node is matched successfully or not needs to integrate the matching results of the node and all child nodes of the node, and therefore, a depth-first traversal strategy needs to be adopted. The depth-first traversal strategy comprises: filtering logic and backtracking logic.
And the filtering logic comprises matching the text with matching information corresponding to the current node, if the matching fails, determining that the matching of the current node fails, if the matching succeeds, checking whether the current node contains an unvisited child node, and if the matching fails, recursively executing the filtering logic on the unvisited child node. That is, the filter logic represents the behavior of the traversal process when it first accesses a node, at which point the traversal process has not accessed any child nodes of that node.
More specifically, the filter logic is operable to determine whether a sufficient requirement for continued access to the child node is that the matching information of the current node matches the text. That is, if the matching information of the current node is successfully matched with the text, inversion can be performed, otherwise, the inverted child node is not required to be performed, and the unmatched child node can be directly determined.
It should be noted that if the current node does not contain an unvisited child node, the backtracking logic is executed on the current node.
Specifically, the backtracking logic includes determining whether all child nodes of the current node fail to match if the current node does not already contain an unvisited child node, if so, the current node is successfully matched, and if at least one child node is successfully matched, the current node is failed to match. That is, the trace-back logic represents the behavior of the traversal process when it last accesses a certain node, and at this point, the traversal process has completed the filtering logic and the trace-back logic of all child nodes of the node, so that the node contains the traversal information of all child nodes at this time.
More specifically, the backtracking logic is configured to integrate matching information of the current node and all child nodes to obtain a matching result of the current node. The method comprises two conditions, wherein one condition is that the matching information of the current node is not matched, and the current node is directly identified as not matched; and the other is matching information of the current node, wherein if all child nodes of the current node are not matched, the result is determined to be matched, otherwise, if the result of one child node is matched, the matching result of the matching information of the current node is reversed, and the result is not matched.
Therefore, there are many semantic matching results corresponding to the output text and the basic matching information of the root node, which are determined according to the matching results of the relevant nodes, and the following examples are given:
in a first example, if it is determined that all child nodes of the root node fail to be matched according to the matching result of the relevant node, the semantic matching corresponding to the output text and the basic matching information of the root node is successful.
In a second example, if it is determined that at least one child node of the root node is successfully matched according to the matching result of the relevant node, semantic matching corresponding to the output text and the basic matching information of the root node fails.
It should be noted that before the filtering logic is executed recursively on the child nodes that have not been accessed, the qualified matching information corresponding to the child nodes that have not been accessed may be characterized by a data structure of a trie tree or a datatrie tree. Therefore, the plurality of child nodes can be processed in parallel, and the matching efficiency is improved.
In summary, in the semantic disambiguation processing method according to the embodiment of the present invention, by acquiring a text to be matched, a depth-first traversal policy is performed on the text starting from basic matching information corresponding to a root node of a semantic disambiguation data structure, and limited matching information corresponding to a relevant node in the semantic disambiguation data structure is matched, where the depth-first traversal policy includes: and the filtering logic and the backtracking logic determine the matching results of all child nodes of the root node according to the matching results of the related nodes and output semantic matching results corresponding to the text and the basic matching information of the root node. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Based on the above embodiment, how to mark the basic matching information as a root node is described below with different scenarios in conjunction with fig. 4, a depth-first traversal strategy is adopted for matching the entire semantic disambiguation data structure for the text, and a semantic matching result of the keyword is returned. The concrete description is as follows:
as a scene implementation, the root node 1 represents a new basic matching information, the basic matching information has three child nodes 2, 7 and 8 representing reverse semantics, the matching information corresponding to the text and the child node 2 is matched, the matching information corresponding to the text and the child node 7 is matched, the matching information corresponding to the text and the child node 8 is matched, and as long as one child node in the child nodes 2, 7 and 8 is successfully matched, the matching failure of the root node 1 is represented; only if the three child nodes 2, 7 and 8 fail to match, the root node 1 is successfully matched.
For example, only if the child node 2 is successfully matched, the child nodes 3 and 6 of the child node 2 need to be further matched, the text is matched with the matching information corresponding to the child node 3, and the text is matched with the matching information corresponding to the child node 6, and as long as one of the child nodes 3 and 6 is successfully matched, the matching failure of the child node 2 is indicated; only if both child nodes 3 and 6 fail to match, the child node 2 is successfully matched. For example, if the child node 6 is successfully matched, the child nodes of the child node 6 need to be further matched, however, as described in fig. 4, if no child node of the child node 6 contains an unvisited child node, the back-tracking logic is executed on the child node 6.
Further, the child node 6 matching successfully indicates that the child node 2 failed to match, and thus, the three child nodes 2, 7, and 8 all failed to match, indicating that the root node 1 was successfully matched. Where the root node 1 and child node 6 are matched.
As another scenario implementation, the root node 1 represents a new basic matching information, the basic matching information has three child nodes 2, 7, and 8 representing inverted semantics, the matching information corresponding to the text and the child node 2 is matched, the matching information corresponding to the text and the child node 7 is matched, and the matching information corresponding to the text and the child node 8 is matched, and as long as one child node of the child nodes 2, 7, and 8 is successfully matched, the matching of the root node 1 is failed; only if the three child nodes 2, 7 and 8 fail to match, the root node 1 is successfully matched.
For example, only if the child node 8 is successfully matched, the child nodes 9 and 12 of the child node 8 need to be further matched, the text is matched with the matching information corresponding to the child node 9, and the text is matched with the matching information corresponding to the child node 12, and as long as one of the child nodes 9 and 12 is successfully matched, the matching failure of the child node 8 is indicated; if both child nodes 9 and 12 fail to match, the child node 8 is successfully matched. For example, the child nodes 9 and 12 fail to match, that is, the child node 8 succeeds in matching, inversion starts to be performed, the child node 8 succeeds in matching and indicates that the root node 1 fails in matching, and the result is directly determined to be an unmatched output result.
In order to make the above scenario more clear to those skilled in the art, the following is exemplified by the basic matching information "golf" and "popular" in conjunction with fig. 5 and 6, respectively:
in a first example, the filter text is "golf car related information" and the basic matching information is determined to be "golf". As shown in fig. 5, the root node "golf" represents a new basic matching information having three sub-nodes "golf", "golf car" and "golf" representing inverted semantics, and matching the text with the matching information corresponding to the sub-node "golf", matching the text with the matching information corresponding to the sub-node "golf car", and matching the text with the matching information corresponding to the sub-node "golf car", it can be seen that the matching of the sub-node "golf car" is successful, indicating that the matching of the root node "golf" is failed.
Further, it is necessary to match the child node "introduction of performance of golf car" of the child node "golf car" with the matching information corresponding to the child node "introduction of performance of golf car" and match the text with the matching information corresponding to the child node "relationship of golf to car", and only if the child node "introduction of performance of golf car" is successfully matched, it is necessary to further match the child node "introduction of performance of golf car", whereas if the child node "introduction of performance of golf car" does not include an unvisited child node, a backtracking logic is performed on the child node "introduction of performance of golf car".
Further, the child node "introduction of performance of golf car" matching success means that the child node "golf car" matching failed, and thus, the child nodes "golf ball", "golf car" and "golf shot" all failed to match, indicating that the root node "golf" matching succeeded. Wherein the root node "golf" and the child node "introduction to performance of golf car" are matched.
In the second example, the screened text is 'authenticity of the popular comment net', and the basic matching information is 'popular'. As shown in fig. 6, the root node "public" represents a new basic matching information, the basic matching information has two sub-nodes "public vehicle" and "public comment" representing inverted semantics, the text is matched with the matching information corresponding to the sub-node "public vehicle", the text is matched with the matching information corresponding to the sub-node "public comment", and if the sub-node "public comment" is successfully matched, the matching of the root node "public" is failed.
Furthermore, the child node 'public vehicle comment condition' of the child node 'public vehicle comment' and 'most audiences comment on a certain film' which need to be subjected to the child node 'public comment' are matched unsuccessfully, that is, the child node 'public vehicle comment condition' and 'most audiences comment on a certain film' both fail to match, means that the child node 'public comment' is matched successfully, the process starts to be reversed, the child node 'public comment' is matched successfully, means that the root node 'public' is matched unsuccessfully, and the child node is directly judged to be unmatched to output a result.
Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Corresponding to the semantic disambiguation processing methods provided in the foregoing embodiments, an embodiment of the present invention further provides a semantic disambiguation processing apparatus, and since the semantic disambiguation processing apparatus provided in the embodiment of the present invention corresponds to the semantic disambiguation processing methods provided in the foregoing embodiments, the embodiments of the semantic disambiguation processing method described above are also applicable to the semantic disambiguation processing apparatus provided in this embodiment, and will not be described in detail in this embodiment.
Fig. 7 is a schematic structural diagram of a semantic disambiguation processing apparatus according to a first embodiment of the present invention, as shown in fig. 7, the semantic disambiguation processing apparatus including: a determination module 11, a generation module 12 and a setup module 13.
And a determining module 11, configured to determine basic matching information of the screened text, where the basic matching information has multiple meanings.
And a generating module 12, configured to generate a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information.
An establishing module 13, configured to establish a semantic disambiguation data structure for screening a text matching the basic matching information and conforming to the target semantic, where the data structure includes: the method comprises the steps of setting a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to a current parent node, and if matching is successful, determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node.
Specifically, in practical applications, there are many forms of basic matching information, and the basic matching information can be selectively set according to practical application needs, for example, as follows:
a first example, a keyword string.
A second example, regular expression.
Further, the generating module 12 generates a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information, and may be selected and set according to actual application requirements, for example, as follows:
in a first example, the qualified matching information is generated according to a preset matching algorithm.
In a second example, qualified match information is generated based on preset stop words.
In a third example, the qualified match information is generated according to a preset context range.
Specifically, a semantic disambiguation data structure for screening semantic text matching the basic matching information and conforming to the target semantic is established. The semantic disambiguation data structures have various forms, and can be selectively set according to actual application needs, for example, as follows:
in a first example, a json structure is used to represent a semantic disambiguation data structure.
The second example, represents the semantic disambiguation data structure using an XML structure.
In summary, the semantic disambiguation processing apparatus according to the embodiment of the present invention determines the basic matching information with multiple meanings of the screened text, generates multiple limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information, and finally establishes a semantic disambiguation data structure for screening the text matching the basic matching information and conforming to the target semantics, where the data structure includes: the method comprises the steps of setting a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and if matching is successful, determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect is achieved.
Based on the embodiment, a complete matching semantic disambiguation data structure is loaded and constructed in the system, basic matching information is marked as a root node, a depth-first traversal strategy is adopted for the text to match the whole semantic disambiguation data structure, and a semantic matching result of the keyword is returned.
Fig. 8 is a schematic structural diagram of a semantic disambiguation processing apparatus according to another embodiment of the present invention, as shown in fig. 8, after the above embodiment, the semantic disambiguation processing apparatus further includes: a matching module 14 and a processing module 15.
The obtaining and matching module 14 is configured to obtain a text to be matched, start from basic matching information corresponding to a root node of a semantic disambiguation data structure, adopt a depth-first traversal policy for the text, and match limited matching information corresponding to related nodes in the semantic disambiguation data structure, where the depth-first traversal policy includes: filtering logic and trace back logic, wherein,
the filtering logic includes: and matching the text with matching information corresponding to the current node, if the matching fails, determining that the matching of the current node fails, if the matching succeeds, checking whether the current node contains the child node which is not accessed, and if the matching succeeds, recursively executing filtering logic on the child node which is not accessed.
The backtracking logic comprises: if the current node does not contain the child nodes which are not visited, judging whether all child nodes of the current node fail to be matched, if so, successfully matching the current node, and if at least one child node succeeds in matching, failing to match the current node.
And the processing module 15 is configured to determine matching results of all child nodes of the root node according to the matching results of the relevant nodes, and output a semantic matching result corresponding to the text and the basic matching information of the root node.
It should be noted that before performing the filtering logic recursively on the child nodes that have not been visited, the limited matching information corresponding to the child nodes that have not been visited may be characterized by the data structure of the trie tree or the datatrie tree, so as to improve the matching efficiency.
Further, in a form in which the present invention may be implemented, the processing module 15 is specifically configured to:
and if all child nodes of the root node are determined to be failed to be matched according to the matching result of the related nodes, the semantic matching between the output text and the corresponding basic matching information of the root node is successful.
And if the matching of at least one child node of the root node is determined to be successful according to the matching result of the related node, the semantic matching corresponding to the basic matching information of the output text and the root node is failed.
In summary, the semantic disambiguation processing apparatus according to the embodiment of the present invention, by acquiring a text to be matched, starting with basic matching information corresponding to a root node of a semantic disambiguation data structure, and adopting a depth-first traversal policy for the text, and matching limited matching information corresponding to a relevant node in the semantic disambiguation data structure, where the depth-first traversal policy includes: and the filtering logic and the backtracking logic determine the matching results of all child nodes of the root node according to the matching results of the related nodes and output semantic matching results corresponding to the text and the basic matching information of the root node. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect can be achieved.
In order to implement the above embodiment, the present invention further provides a server. FIG. 9 is an interaction flow diagram of a semantic disambiguation processing method according to an embodiment of the invention, wherein the semantic disambiguation processing on the server side comprises the following steps: the processor firstly determines basic matching information of the screened text, wherein the basic matching information has multiple meanings, then the processor generates a plurality of limited matching information associated with the basic matching information according to target semantics and ambiguous semantics corresponding to the basic matching information, and finally the processor establishes a semantic disambiguation data structure which is used for screening the text matched with the basic matching information and conforms to the target semantics, wherein the data structure comprises: the method comprises the steps of generating basic matching information, generating a plurality of limiting matching information, generating a root node corresponding to the basic matching information and a plurality of child nodes corresponding to the limiting matching information, and establishing a relation set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, wherein the matching result of the parent node and a text is reversed when the matching information corresponding to the child nodes is successfully matched with the text.
To sum up, the server according to the embodiment of the present invention determines the multi-semantic basic matching information of the screened text, generates a plurality of limited matching information associated with the basic matching information according to the target semantic meaning and the ambiguous semantic meaning corresponding to the basic matching information, and finally establishes a semantic disambiguation data structure for screening the text matching the basic matching information and conforming to the target semantic meaning, where the data structure includes: the method comprises the steps of setting a root node corresponding to basic matching information and a plurality of child nodes corresponding to a plurality of limited matching information, establishing a relationship set of a parent node and the child nodes corresponding to the root node and the child nodes according to target semantics and ambiguous semantics, matching a text to be matched with matching information corresponding to the current parent node, and determining whether to reverse the matching result with the parent node according to the matching result with the corresponding child node when the matching is successful. Therefore, the efficiency and the accuracy of semantic disambiguation can be improved, and a better semantic matching effect can be achieved.
In order to implement the foregoing embodiments, the present invention further provides a storage medium for storing an application program, where the application program is configured to execute the semantic disambiguation processing method according to any one of the embodiments of the present invention.
In the description of the present invention, it is to be understood that the terms "first", "second", and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or to imply that the number of technical features indicated are in fact significant. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.
In the description of the specification, reference to the description of the term "one embodiment", "some embodiments", "an example", "a specific example", or "some examples", etc., means that a particular feature or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Moreover, various embodiments or examples and features of various embodiments or examples described in this specification can be combined and combined by one skilled in the art without being mutually inconsistent.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps in the process, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present invention.
The logic and/or steps represented in the flowcharts or otherwise described herein, such as an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a separate product, may also be stored in a computer-readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (12)
1. A semantic disambiguation processing method, comprising:
determining basic matching information of the screened text, wherein the basic matching information has multiple meanings;
generating a plurality of limited matching information associated with the basic matching information according to the target semantics and the ambiguous semantics corresponding to the basic matching information;
establishing a semantic disambiguation data structure for screening matching the basic matching information and conforming to a target semantic text, the data structure comprising: and establishing a relationship set of a parent node and child nodes corresponding to the root node and the child nodes according to the target semantics and the ambiguous semantics, wherein the text to be matched is matched with the matching information corresponding to the current parent node, and if the matching is successful, whether the matching result with the parent node is reversed is determined according to the matching result with the corresponding child node.
2. The method of claim 1, wherein the generating a plurality of qualified match information associated with the base match information comprises:
generating the limited matching information according to a preset matching algorithm; or,
generating the limited matching information according to a preset stop word; or,
and generating the limited matching information according to a preset context range.
3. The method of claim 1 or 2, further comprising:
acquiring a text to be matched, starting from basic matching information corresponding to a root node of the semantic disambiguation data structure, adopting a depth-first traversal strategy for the text, and matching limited matching information corresponding to related nodes in the semantic disambiguation data structure, wherein the depth-first traversal strategy comprises the following steps: filtering logic and trace back logic, wherein,
the filtering logic includes: matching the text with matching information corresponding to the current node, if the matching fails, determining that the matching of the current node fails, if the matching succeeds, checking whether the current node contains child nodes which are not accessed, and if so, recursively executing the filtering logic on the child nodes which are not accessed;
the backtracking logic comprises: if the current node does not contain the child nodes which are not visited, judging whether all child nodes of the current node fail to be matched, if so, successfully matching the current node, and if at least one child node succeeds in matching, failing to match the current node;
and determining the matching results of all child nodes of the root node according to the matching results of the related nodes, and outputting semantic matching results corresponding to the text and the basic matching information of the root node.
4. The method of claim 3, wherein prior to recursively executing the filter logic on the unvisited child nodes, further comprising:
and characterizing the limited matching information corresponding to the child nodes which are not visited through a data structure of a trie tree or a datatrie tree.
5. The method of claim 3, wherein determining matching results for all children nodes of the root node based on the matching results for the relevant nodes and outputting semantic matching results for the text corresponding to the base matching information for the root node comprises:
if all child nodes of the root node are determined to be failed to be matched according to the matching result of the related node, outputting the semantic matching success corresponding to the text and the basic matching information of the root node;
and if the matching of at least one child node of the root node is determined to be successful according to the matching result of the related node, outputting semantic matching failure corresponding to the text and the basic matching information of the root node.
6. A semantic disambiguation processing apparatus, comprising:
the determination module is used for determining basic matching information of the screened text, wherein the basic matching information has multiple meanings;
a generating module, configured to generate multiple pieces of limited matching information associated with the basic matching information according to a target semantic and an ambiguous semantic corresponding to the basic matching information;
an establishing module for establishing a semantic disambiguation data structure for screening semantic texts matching the basic matching information and conforming to a target, the data structure comprising: and establishing a relationship set of a parent node and child nodes corresponding to the root node and the child nodes according to the target semantics and the ambiguous semantics, wherein if the text to be matched is matched with the matching information corresponding to the current parent node, if the matching is successful, whether the matching result with the parent node is reversed is determined according to the matching result with the corresponding child node.
7. The apparatus of claim 6, wherein the generation module is specifically configured to:
generating the limited matching information according to a preset matching algorithm; or,
generating the limited matching information according to a preset stop word; or,
and generating the limited matching information according to a preset context range.
8. The apparatus of claim 6 or 7, further comprising:
the acquisition matching module is used for acquiring a text to be matched, starting from basic matching information corresponding to a root node of the semantic disambiguation data structure, adopting a depth-first traversal strategy for the text, and matching limited matching information corresponding to related nodes in the semantic disambiguation data structure, wherein the depth-first traversal strategy comprises the following steps: filtering logic and trace back logic, wherein,
the filtering logic comprises: matching the text with matching information corresponding to the current node, if the matching fails, determining that the matching of the current node fails, if the matching succeeds, checking whether the current node contains child nodes which are not accessed, and if so, recursively executing the filtering logic on the child nodes which are not accessed;
the backtracking logic comprises: if the current node does not contain the child nodes which are not visited, judging whether all child nodes of the current node fail to be matched, if so, successfully matching the current node, and if at least one child node succeeds in matching, failing to match the current node;
and the processing module is used for determining the matching results of all child nodes of the root node according to the matching results of the related nodes and outputting semantic matching results corresponding to the text and the basic matching information of the root node.
9. The apparatus of claim 8, wherein prior to recursively executing the filter logic on the unvisited child nodes, further comprising:
and characterizing the limited matching information corresponding to the child node which is not visited through a data structure of a trie tree or a datatrie tree.
10. The apparatus of claim 8, wherein the processing module is specifically configured to:
if all child nodes of the root node are determined to be failed to be matched according to the matching result of the related node, outputting the semantic matching success corresponding to the text and the basic matching information of the root node;
and if the matching of at least one child node of the root node is determined to be successful according to the matching result of the related node, outputting semantic matching failure corresponding to the text and the basic matching information of the root node.
11. A server, comprising:
memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor implements the semantic disambiguation processing method according to any of the claims 1-5 when executing the computer program.
12. A storage medium for storing an application program for executing the semantic disambiguation processing method of any one of claims 1 through 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710585495.2A CN109271621B (en) | 2017-07-18 | 2017-07-18 | Semantic disambiguation processing method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710585495.2A CN109271621B (en) | 2017-07-18 | 2017-07-18 | Semantic disambiguation processing method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109271621A CN109271621A (en) | 2019-01-25 |
| CN109271621B true CN109271621B (en) | 2023-04-18 |
Family
ID=65147920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710585495.2A Active CN109271621B (en) | 2017-07-18 | 2017-07-18 | Semantic disambiguation processing method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109271621B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11985023B2 (en) | 2018-09-27 | 2024-05-14 | Juniper Networks, Inc. | Supporting graphQL based queries on yang based configuration data models |
| US10892952B2 (en) | 2019-02-21 | 2021-01-12 | Juniper Networks, Inc. | Supporting compilation and extensibility on unified graph-based intent models |
| US10841182B2 (en) | 2019-03-29 | 2020-11-17 | Juniper Networks, Inc. | Supporting near real time service level agreements |
| US10897396B2 (en) | 2019-03-29 | 2021-01-19 | Juniper Networks, Inc. | Supporting concurrency for graph-based high level configuration models |
| EP3722944A1 (en) | 2019-04-10 | 2020-10-14 | Juniper Networks, Inc. | Intent-based, network-aware network device software-upgrade scheduling |
| US11165647B2 (en) * | 2019-06-28 | 2021-11-02 | Juniper Networks, Inc. | Managing multiple semantic versions of device configuration schemas |
| CN110703614B (en) * | 2019-09-11 | 2021-01-22 | 珠海格力电器股份有限公司 | Voice control method and device, semantic network construction method and device |
| CN112379961A (en) * | 2020-11-23 | 2021-02-19 | 厦门市美亚柏科信息股份有限公司 | Method for representing multiple semantics by mesh topology structure, terminal equipment and storage medium |
| CN112784046B (en) * | 2021-01-20 | 2024-05-28 | 北京百度网讯科技有限公司 | Text clustering method, device, equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5317507A (en) * | 1990-11-07 | 1994-05-31 | Gallant Stephen I | Method for document retrieval and for word sense disambiguation using neural networks |
| EP1557824A1 (en) * | 2004-01-22 | 2005-07-27 | AT&T Corp. | System and method to disambiguate user's intention in a spoken dialog system |
| CN1871597A (en) * | 2003-08-21 | 2006-11-29 | 伊迪利亚公司 | System and method for associating documents with contextual advertisements |
| CN1916887A (en) * | 2006-09-06 | 2007-02-21 | 哈尔滨工程大学 | Method for eliminating ambiguity without directive word meaning based on technique of substitution words |
| CN103577491A (en) * | 2012-08-09 | 2014-02-12 | 佳能株式会社 | Method and device for representing functional entities and carrying out disambiguation on functional entities |
| CN104142917A (en) * | 2014-05-21 | 2014-11-12 | 北京师范大学 | Hierarchical semantic tree construction method and system for language understanding |
| CN104182420A (en) * | 2013-05-27 | 2014-12-03 | 华东师范大学 | Ontology-based Chinese name disambiguation method |
| CN105893346A (en) * | 2016-03-30 | 2016-08-24 | 齐鲁工业大学 | Graph model word sense disambiguation method based on dependency syntax tree |
| CN106709011A (en) * | 2016-12-26 | 2017-05-24 | 武汉大学 | Positional concept hierarchy disambiguation calculation method based on spatial locating cluster |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7711672B2 (en) * | 1998-05-28 | 2010-05-04 | Lawrence Au | Semantic network methods to disambiguate natural language meaning |
| US7899666B2 (en) * | 2007-05-04 | 2011-03-01 | Expert System S.P.A. | Method and system for automatically extracting relations between concepts included in text |
| EP2368201A4 (en) * | 2008-12-09 | 2013-08-07 | Univ Houston System | Word sense disambiguation |
| US20110119047A1 (en) * | 2009-11-19 | 2011-05-19 | Tatu Ylonen Oy Ltd | Joint disambiguation of the meaning of a natural language expression |
| RU2592395C2 (en) * | 2013-12-19 | 2016-07-20 | Общество с ограниченной ответственностью "Аби ИнфоПоиск" | Resolution semantic ambiguity by statistical analysis |
| US20160335367A1 (en) * | 2015-05-15 | 2016-11-17 | Microsoft Technology Licensing, Llc | Entity disambiguation using multisource learning |
-
2017
- 2017-07-18 CN CN201710585495.2A patent/CN109271621B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5317507A (en) * | 1990-11-07 | 1994-05-31 | Gallant Stephen I | Method for document retrieval and for word sense disambiguation using neural networks |
| CN1871597A (en) * | 2003-08-21 | 2006-11-29 | 伊迪利亚公司 | System and method for associating documents with contextual advertisements |
| EP1557824A1 (en) * | 2004-01-22 | 2005-07-27 | AT&T Corp. | System and method to disambiguate user's intention in a spoken dialog system |
| CN1916887A (en) * | 2006-09-06 | 2007-02-21 | 哈尔滨工程大学 | Method for eliminating ambiguity without directive word meaning based on technique of substitution words |
| CN103577491A (en) * | 2012-08-09 | 2014-02-12 | 佳能株式会社 | Method and device for representing functional entities and carrying out disambiguation on functional entities |
| CN104182420A (en) * | 2013-05-27 | 2014-12-03 | 华东师范大学 | Ontology-based Chinese name disambiguation method |
| CN104142917A (en) * | 2014-05-21 | 2014-11-12 | 北京师范大学 | Hierarchical semantic tree construction method and system for language understanding |
| CN105893346A (en) * | 2016-03-30 | 2016-08-24 | 齐鲁工业大学 | Graph model word sense disambiguation method based on dependency syntax tree |
| CN106709011A (en) * | 2016-12-26 | 2017-05-24 | 武汉大学 | Positional concept hierarchy disambiguation calculation method based on spatial locating cluster |
Non-Patent Citations (9)
| Title |
|---|
| Andrei Minca et al.An approach to knowledge-based Word Sense Disambiguation using semantic trees built on a WordNet lexicon network.《2011 6th Conference on Speech Technology and Human-Computer Dialogue》.2011,1-6. * |
| Federica Mandreoli et al.Versatile Structural Disambiguation for Semantic-aware Applications.《Proceedings of the 2005 ACM CIKM International Conference on Information and Knowledge》 .2005,1-8. * |
| Patrick Ye et al .Verb Sense Disambiguation Using Selectional Preferences Extracted with a State-of-the-art Semantic Role Labeler.《Proceedings of the 2006 Australasian Language Technology Workshop》.2006,139-148. * |
| Ping Chen et al .A Fully Unsupervised Word Sense Disambiguation Method Using Dependency Knowledge.《The 2009 Annual Conference of the North American Chapter of the ACL》 .2009,28-36. * |
| 刘小宇.基于语义理解的中文常问问答系统的研究.《中国优秀博硕士学位论文全文数据库 (硕士)信息科技辑》.2007,I138-99. * |
| 刘运通 等.基于动态规划的简单语义单元词义消歧.《计算机工程与设计》.2014,第35卷(第4期),1480-1485. * |
| 孙红俊 等 .业务过程执行语言流程语义标注方法研究.《计算机集成制造系统》.2009,第15卷(第3期),609-617. * |
| 左乃彻.基于维基百科的中英文命名实体消歧.《中国优秀硕士学位论文全文数据库信息科技辑》.2015,I138-1580. * |
| 罗俊丽 等 .基于词义消歧的语义查询扩展研究.《微电子学与计算机》.2012,第29卷(第1期),71-75. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109271621A (en) | 2019-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109271621B (en) | Semantic disambiguation processing method, device and equipment | |
| EP3678346B1 (en) | Blockchain smart contract verification method and apparatus, and storage medium | |
| CN108089893B (en) | Method and device for determining redundant resources, terminal equipment and storage medium | |
| KR20180041200A (en) | Information processing method and apparatus | |
| CN106157979B (en) | A kind of method and apparatus obtaining voice pitch data | |
| CN109558525A (en) | A kind of generation method of test data set, device, equipment and storage medium | |
| CN111563192A (en) | Entity alignment method and device, electronic equipment and storage medium | |
| CN113672389B (en) | Server compatible method, system, equipment and computer readable storage medium | |
| CN105787126A (en) | K-d (k-dimensional) tree generation method and k-d tree generation device | |
| CN111831626A (en) | Graph structure generation method of database logical relation, data query method and device | |
| Voigt et al. | Matchbox: combined meta-model matching for semi-automatic mapping generation | |
| Bollig et al. | Parameterized verification of communicating automata under context bounds | |
| CN105843809A (en) | Data processing method and device | |
| CN112052157A (en) | Test message construction method, device and system | |
| CN114385764A (en) | Reverse analysis method and device for relational database table structure | |
| CN113064898A (en) | Retrieval method and device based on miniature index of contract on chain and electronic equipment | |
| US10044594B2 (en) | Collaboration method and device for verifying behavior in route switching system | |
| CN103440339A (en) | Method, device and system for searching application system change association objects | |
| CN112163430B (en) | Method and system for ontology fusion of industrial equipment data facing MES system | |
| CN114281830B (en) | Rule mapping table construction method, rule matching method and device for multi-attribute conditions | |
| CN118194862B (en) | Text error correction method and device based on fault-tolerant suffix automaton | |
| CN115047835B (en) | DCS-based periodic test data acquisition method, device, equipment and medium | |
| JP2013148953A (en) | Program template automatic generation program and program template automatic generation device | |
| Wiedermann et al. | Computing by Broadcasting: Alternation in Disguise (Extended Version) | |
| CN118642756A (en) | Tree structure-based instruction decoding conflict detection method, instruction decoding method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |