CN110096692A - A kind of Semantic Information Processing method and apparatus - Google Patents

A kind of Semantic Information Processing method and apparatus Download PDF

Info

Publication number
CN110096692A
CN110096692A CN201810081517.6A CN201810081517A CN110096692A CN 110096692 A CN110096692 A CN 110096692A CN 201810081517 A CN201810081517 A CN 201810081517A CN 110096692 A CN110096692 A CN 110096692A
Authority
CN
China
Prior art keywords
semantic information
conclusion
conditions
dominant
latent
Prior art date
Application number
CN201810081517.6A
Other languages
Chinese (zh)
Inventor
刘飞飞
Original Assignee
亿度慧达教育科技(北京)有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 亿度慧达教育科技(北京)有限公司 filed Critical 亿度慧达教育科技(北京)有限公司
Priority to CN201810081517.6A priority Critical patent/CN110096692A/en
Publication of CN110096692A publication Critical patent/CN110096692A/en

Links

Classifications

    • G06F17/278
    • G06F17/2785

Abstract

The embodiment of the invention provides a kind of Semantic Information Processing method and apparatus, belong to technical field of information processing.The Semantic Information Processing method includes that stem is divided into known conditions and conclusion two parts according to obtained stem;According to obtained known conditions and conclusion, the dominant semantic information in known conditions and conclusion is extracted;When there are Latent Semantic information when Latent Semantic information, in extraction known conditions and/conclusion in known conditions and/or conclusion;Merge the dominant semantic information and Latent Semantic information extracted, obtains the semantic information of stem.Semantic Information Processing method provided in an embodiment of the present invention can extract more comprehensive, complete semantic information from stem, so as to provide accurate unified input information for the automatic answer of topic.

Description

A kind of Semantic Information Processing method and apparatus

Technical field

The present embodiments relate to the technical field of information processing more particularly to a kind of Semantic Information Processing method and dresses It sets.

Background technique

In order to improve the efficiency of information processing, information required for comprehensively grasping, feature extraction becomes research heat gradually Point.

Semantic-based feature extracting method is one of common feature extracting method, how from a variety of form of presentation Sentence in extract useful information as a urgent problem to be solved.

Summary of the invention

In view of this, one of the technical issues of embodiment of the present invention is solved is to provide a kind of Semantic Information Processing method And device, semantic information can be accurately extracted from the stem of topic, so that the automatic answer for topic provides accurate unification Input.

In a first aspect, the embodiment of the present invention provides a kind of Semantic Information Processing method, comprising:

According to obtained stem, the stem is divided into known conditions and conclusion two parts;

According to obtained known conditions and conclusion, the dominant semantic information in the known conditions and conclusion is extracted;

When, there are when Latent Semantic information, extracting the known conditions and/or conclusion in the known conditions and/or conclusion In Latent Semantic information;

Merge the dominant semantic information and Latent Semantic information extracted, obtains the semantic information of the stem.

Optionally, in the specific embodiment based on first aspect, there are Latent Semantics for the known conditions and/or conclusion The judgment step of information are as follows:

According to the known conditions and/or conclusion, according to the obtained semantic information of the extracting mode of dominant semantic information In when having entity missing, then judging the known conditions and/or conclusion, there are Latent Semantic information.Optionally, the basis obtains The known conditions and conclusion arrived, the step of extracting the dominant semantic information in the known conditions and conclusion specifically:

According to the correspondence table of keyword and relationship, the corresponding relationship of keyword in the known conditions and conclusion is obtained;

According to the positional relationship between the entity in the keyword and the known conditions and conclusion, in conjunction with the key The corresponding relationship of word extracts the dominant semantic information in the known conditions and conclusion.

Optionally, when the known conditions and/or conclusion there are when Latent Semantic information, extract the known conditions and/ Or the specific steps of the Latent Semantic information in conclusion:

When having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information, lacked according to existing Relationship in the semantic information of unfounded body obtains the corresponding keyword of the relationship;

According to the sentence before or after the keyword, the recessive language in the known conditions and/or conclusion is extracted Adopted information.

Optionally, the known conditions and conclusion that the basis obtains, extract the dominant language in the known conditions and conclusion The step of adopted information specifically:

According to the comma in the known conditions and conclusion, the known conditions divided by the comma and conclusion are obtained In single statement;

Dominant semantic information is extracted from the single statement respectively;

Merge the dominant semantic information extracted from each single statement, obtains in the known conditions and conclusion Dominant semantic information.

Optionally, described to merge the dominant semantic information and Latent Semantic information extracted, obtain the semanteme of the stem Before information further include:

When extracting the known conditions and/or knot according to the sentence before or after the sentence where the keyword During Latent Semantic information in, if having found the supplement semantic information of the stem, the supplement language is extracted Adopted information, and the supplement semantic information is denoted as Latent Semantic information.

Second aspect, the embodiment of the invention also provides a kind of Semantic Information Processing devices, comprising:

Division module, for according to obtained stem, the stem to be divided into known conditions and conclusion two parts;

Dominant Semantic features extraction module, for according to obtained known conditions and conclusion, extract the known conditions and Dominant semantic information in conclusion;

Latent Semantic information extraction modules, for mentioning when the known conditions and/or conclusion are there are when Latent Semantic information Take the Latent Semantic information in the known conditions and/or conclusion;

Merging module, for merging dominant semantic information that the dominant Semantic features extraction module is extracted and described hidden The Latent Semantic information that property Semantic features extraction module is extracted, obtains the semantic information of the stem.Optionally, based on the In the specific embodiment of the invention of two aspects, described device further includes judgment module, is used for:

According to the known conditions and/or conclusion, according to the obtained semantic information of the extracting mode of dominant semantic information In when having entity missing, then judging the known conditions and/or conclusion, there are Latent Semantic information;

The Latent Semantic information extraction modules are connected with the judgment module, described for working as judgment module judgement Know that there are Latent Semantic when Latent Semantic information, extracted in the known conditions and/or conclusion letters in condition and/or conclusion Breath;

The merging module is specifically used for merging the dominant semantic information that the dominant Semantic features extraction module is extracted The Latent Semantic information extracted with the Latent Semantic information extraction modules, obtains the semantic information of the stem.

Optionally, the dominant Semantic features extraction module is specifically used for: according to the correspondence table of keyword and relationship, obtaining The corresponding relationship of keyword in the known conditions and conclusion;

According to the positional relationship between the entity in keyword and the known conditions and conclusion, in conjunction with the keyword pair The relationship answered extracts the dominant semantic information in the known conditions and conclusion.

Optionally, the Latent Semantic information extraction modules are specifically used for:

It is real according to existing when having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information Relationship in the semantic information of body missing, obtains the corresponding keyword of the relationship;

According to the sentence before or after keyword, the Latent Semantic letter in the known conditions and/or conclusion is extracted Breath, and,

When extracting the known conditions and/or knot according to the sentence before or after the sentence where the keyword During Latent Semantic information in, if having found the supplement semantic information of the stem, the supplement is extracted Semantic information, and the supplement semantic information is denoted as Latent Semantic information.

By above technical scheme as it can be seen that the embodiment of the present invention can extract comprehensive stem information from stem, extract Stem semantic information group out can include information all in stem, thus for topic it is automatic answer provide it is accurate and effective And unified input.

Detailed description of the invention

In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this The some embodiments recorded in inventive embodiments can also obtain according to these attached drawings for those of ordinary skill in the art Obtain other attached drawings.

Fig. 1 is a kind of flow diagram of Semantic Information Processing method provided in an embodiment of the present invention;

Fig. 2 is a kind of structure drawing of device of Semantic Information Processing device provided in an embodiment of the present invention.

Specific embodiment

In order to make those skilled in the art more fully understand the technical solution in the embodiment of the present invention, below in conjunction with the present invention Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described reality Applying example only is a part of the embodiment of the embodiment of the present invention, instead of all the embodiments.Based on the implementation in the embodiment of the present invention The range of protection of the embodiment of the present invention all should belong in example, those of ordinary skill in the art's every other embodiment obtained.

Referring to Fig. 1, the embodiment of the invention provides a kind of semantic processes method, including step S100-S400, specifically Are as follows:

S100: according to obtained stem, the stem is divided into known conditions and conclusion two parts.

It should be noted that Semantic Information Processing method provided by the embodiments of the present application, is applicable to a variety of topic types, including Geometric proof topic, calculation question etc., at this by extract geometric proof topic stem in semantic information for the embodiment of the present application into Row explanation.

The embodiment of the present application finds that the stem of proof question can be used by the statistical analysis to a large amount of proof question stem texts " solving " perhaps " proof " etc. is divided into the part stem before known conditions and conclusion two parts " solving " or " proof " is Know condition, the part stem after " solving " or " proof " is conclusion.Therefore, in the embodiment of the present application, can be inscribed by traversal It is dry, " solving " or " proof " is found, and then be proven the known conditions and conclusion of topic stem.

It equally, can also be by " calculating ", " solution " etc. to known to stem when extracting the stem semantic information of calculation question Condition and conclusion are classified.

It should be noted that " semantic information " in the embodiment of the present application includes dominant semantic information and Latent Semantic letter Breath.Wherein, dominant semantic information refers to not needing from the context, according to single statement can immediately arrive at comprising two The semantic information of relationship between entity and two entities, described two entities can be respectively distributed to corresponding relationship Two sides, and in every side, the entity is most short at a distance from corresponding relationship, such as AB ⊥ CD;Can also be distributed in pair The same side for the relationship answered, if AC is the diagonal line of diamond shape ABCD;Latent Semantic information refers to needing from the context just may be used With the information obtained, for geometric proof topic, " semantic information " refers to the mathematical information for including in stem sentence.

S200: according to obtained known conditions and conclusion, the dominant semantic information in the known conditions and conclusion is extracted.

In actual operation, optionally, it according to the correspondence table of keyword and relationship, obtains in the known conditions and conclusion The corresponding relationship of keyword;According to the positional relationship between the entity in the keyword and the known conditions and conclusion, In conjunction with the corresponding relationship of the keyword, the dominant semantic information in the known conditions and conclusion is extracted.

It should be noted that the entity in the embodiment of the present application refers to the specific unit of basic conception in mathematics, such as three Angular ABC, line segment AB, point O etc.;Relationship refers to the relationship between entity and entity, such as vertical, conllinear etc..Semantic information refers to The set being made of entity and relationship, can be indicated with { entity 1, entity 2, the relationship of entity 1 and entity 2 }, for example, topic There is the sentence of AB ⊥ CD in dry known conditions or conclusion, can obtain the corresponding semantic information of this sentence is that { AB, CD hang down Directly }.

Optionally, according to the comma in the known conditions and conclusion, the known item divided by the comma is obtained Single statement in part and conclusion;Dominant semantic information is extracted from the single statement respectively;Merge from each described single The dominant semantic information extracted in sentence obtains the dominant semantic information in the known conditions and conclusion.

It, can be according to the single statement in the known conditions and/or conclusion divided with comma, from this in practical implementation The keyword that can embody relationship between entity is searched in a little single statements, the foundation for searching keyword can refer to table 1:

1 keyword of table and relationship correspond to table

It is proven after the corresponding relationship of keyword in topic stem, according to the position of relationship and entity in sentence Relationship extracts dominant semantic information.For example, if having a single statement in the known conditions of proof question stem is " DE ⊥ AB ", then it is right to obtain keyword " ⊥ " by searching for table 1 by traversal sentence " DE ⊥ AB " available keyword " ⊥ " The relationship answered is vertical, and respectively has an entity in the two sides of keyword " ⊥ ", is " DE " and " AB " respectively, therefore according to key The positional relationship of word " ⊥ " and entity in sentence " DE ⊥ AB " can extract the corresponding dominant semantic letter of sentence " DE ⊥ AB " Breath { DE, AB, vertical } or { AB, DE, vertical }.

S300: when in the known conditions and/or conclusion there are when Latent Semantic information, extract the known conditions and/ Or the Latent Semantic information in conclusion.

During the semantic information of actual extracting stem, by the way that after finding relation keyword, search relationship is crucial The dominant semantic information that the entity of word two sides or the entity of keyword the same side obtain be possible to be not stem complete language The case where adopted information, i.e., there is likely to be Latent Semantic information in the known conditions of stem and/or conclusion.

Optionally, according to the known conditions and/or conclusion, according to the obtained language of the extracting mode of dominant semantic information When having entity missing in adopted information, then judging the known conditions and/or conclusion, there are Latent Semantic information.

Further, when having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information, According to the relationship in the semantic information that there is missing entity, the corresponding keyword of the relationship is obtained;According to the keyword it Preceding and/or sentence later, extract the known conditions and/or in Latent Semantic information.

For example, if obtained, " as schemed, diagonal line AC, BD of diamond shape ABCD intersects at point O, BE ∥ AC, CE ∥ DB. solve: the stem of BE ⊥ CE. ", in the embodiment of the present application, first according to " solving " by stem be divided into known conditions and Conclusion two parts:

Known conditions: as schemed, diagonal line AC, BD of diamond shape ABCD intersects at point O, BE ∥ AC, CE ∥ DB.

Conclusion: BE ⊥ CE.

According to the comma in above-mentioned known conditions and conclusion, the list in the known conditions separated by comma is respectively obtained A sentence:

Single statement a: as schemed.

Single statement b: diamond shape ABCD diagonal line AC, BD intersects at point O.

Single statement c:BE ∥ AC.

Single statement d:CE ∥ DB.

And the single statement in conclusion:

Single statement e:BE ⊥ CE.

Due to from known conditions extract semantic information with from conclusion extract semantic information principle be as, The embodiment of the present application only to from known conditions extract semantic information make elaboration.

For single statement a, by traversing the sentence, finding the sentence and not including the keyword that can embody relationship, Therefore semantic information can not be obtained from single statement a.

Diagonal line is obtained, and intersect at two to embody relationship by traversing the sentence for single statement b Keyword, and correspond to three entities certainly from intersecting at this keyword and can obtain the collinear relationship, two by entity and The semantic set of relationship composition, and diagonal line might not be obtained from this keyword of diagonal line corresponding to three entities, two A semantic set being made of entity and relationship, because being possible to only use a diagonal line in stem.According to keyword With the correspondence table of relationship, i.e., table 1 in the embodiment of the present application, obtaining the corresponding relationship of diagonal line is diagonal line, intersects at correspondence Relationship be conllinear.

Further, because diamond shape ABCD and diagonal line AC is located at the cornerwise two sides of keyword, according to diagonal line two sides And the shortest closed diamonds ABCD and AC of distance between diagonal line, obtain dominant semantic information diamond shape ABCD, AC, diagonally Line };Also according to intersecting at two sides, and with intersect between distance shortest entity B D, point O, and from intersecting at institute itself What is obtained corresponds to three entities, the information of two semantic set being made of entity and relationship, according to dominant semantic information The obtained semantic information of extracting mode be { BD, point O, collinearly } and?, point O, collinearly }, wherein? indicate lacked reality Body.It can estimate at this time, there are Latent Semantic information in single statement b, need to find lacked reality by from the context Body, and the entity lacked is added in corresponding semantic information.

From semantic information?, point O is conllinear } in the relationship that can obtain in the semantic information be conllinear, and collinearly corresponding topic Keyword in dry is to intersect at, to intersect at as the starting point for searching lacked entity, before sentence for " diamond shape ABCD's Diagonal line AC, BD ", sentence thereafter are " point O ", it is evident that do not include lacked entity in sentence thereafter, therefore, with It intersects at as starting point, is searched in the sentence before intersecting at according to direction from right to left, first look for BD entity, this When, entity whether can be missing to BD entity and judged, can also directly add to BD as the entity lacked in the presence of scarce In the semantic information of unfounded body.

On the one hand, in practical implementation, judge if whether to be missing from entity to BD entity, it can be by true Determine whether BD is to be located to intersect at wherein side and distance intersects at nearest entity and judged, if so, BD is not to lack The entity of mistake, if it is not, then can first be added to BD as the entity lacked in the semantic information in the presence of missing entity.

On the other hand, in practical implementation, if do not judged whether BD is missing from entity, can directly by BD is added in the semantic information in the presence of missing entity as the entity lacked.

For above-mentioned two aspect, when in the semantic information for adding to presence missing entity using BD as the entity lacked Afterwards, check the semantic information after supplement is complete whether and the semantic information that has extracted repeat, if repeating, it is concluded that BD is not institute Lack the conclusion of entity;If not repeating, using BD as the entity lacked, it is obvious that in the present embodiment, BD is not lacked Unfounded body.

Based on the above, continue Look-ahead, find ", " before, other than entity B D, distance intersects at nearest entity AC, according to the judgment mode of the above missing entity, AC is added to?, point O, collinearly } in, obtain { AC, point O, conllinear }.

Optionally, when extracting the known conditions according to the sentence before or after the sentence where the keyword And/or during the Latent Semantic information in conclusion, if having found the supplement semantic information of the stem, described in extraction Semantic information is supplemented, and the supplement semantic information is denoted as Latent Semantic information;

By all Latent Semantic information and the dominant merging extracted the dominant semantic information, Latent Semantic information With supplement semantic information, the semantic information of the stem is obtained.

As can be seen from the above embodiments, search lack entity during, it was found that ", ", be located at the pause mark two sides and with The pause mark is respectively AC, BD apart from nearest entity, and the meaning according to representated by pause mark, i.e. pause mark are used to divide phrase arranged side by side, The corresponding relationship of AC and the corresponding relationship consistency of BD are obtained, according to semantic information { diamond shape ABCD, AC, diagonal line }, moreover it is possible to obtain { diamond shape ABCD, BD, diagonal line } such a supplement semantic information.

Certainly, the embodiment of the present application can also by pause mark ", ", be distributed in pause mark ", " the entity A C, BD, Yi Jiyi of two sides The semantic information { BD, point O, conllinear } extracted obtains the Latent Semantic information of { AC, point O, conllinear }.

S400: merging the semantic information extracted, and obtains the semantic information of the stem.

In conclusion according to single statement b, the corresponding dominant semantic information of available single statement b is { diamond shape ABCD, AC, diagonal line }, { BD, point O, conllinear } and Latent Semantic information are { AC, point O, collinearly }, diamond shape ABCD, BD, it is right Linea angulata }, above-mentioned dominant semantic information and Latent Semantic information are merged, obtaining the corresponding semantic information of single statement b is { water chestnut Shape ABCD, AC, diagonal line }, { BD, point O, conllinear }, { AC, point O, conllinear } and { diamond shape ABCD, BD, diagonal line }.

Equally, corresponding semantic information can be obtained according to single statement c is { BE, AC, parallel }, according to single statement It is { CE, DB, parallel } that d, which obtains corresponding semantic information, and obtaining corresponding semantic information according to single statement e is { BE, CE, vertical }.

From the foregoing, it will be observed that obtaining the corresponding semantic information of known conditions in the stem is { diamond shape ABCD, AC, diagonal line }, { diamond shape ABCD, BD, diagonal line }, { BD, point O, conllinear } and { AC, point O, conllinear }, { BE, AC, parallel }, { CE, DB are put down Row }, { BE, CE, vertical }.

According to method same with the semantic information of known conditions in stem is extracted, the corresponding language of conclusion BE ⊥ CE can be obtained Adopted information is { BE, CE, vertical }.Therefore the corresponding semantic information of known conditions and the corresponding semantic information of conclusion are merged, it can Obtain the semantic information of the stem are as follows:

Known conditions: { diamond shape ABCD, AC, diagonal line }, { diamond shape ABCD, BD, diagonal line }, { BD, point O, conllinear } and { AC, point O, conllinear }, { BE, AC, parallel }, { CE, DB, parallel };

Conclusion: { BE, CE, vertical }.

When concrete application, the semantic information of the stem can be input in automated answering question system, so as to automated answering question system Topic can voluntarily be answered according to the information of input.

A kind of Semantic Information Processing method provided by the embodiments of the present application, there are when Latent Semantic information in stem, Recessive Semantic features extraction can be come out, can be realized entity and relationship all standing, it is comprehensive and complete so as to extract Semantic information, provide accurate unified input information for the automatic answer of topic.

Based on identical inventive concept, referring to Fig.2, the embodiment of the present application also provides a kind of Semantic Information Processing device, Include:

Division module 201, for according to obtained stem, the stem to be divided into known conditions and conclusion two parts;

Dominant Semantic features extraction module 202, for extracting the known conditions according to obtained known conditions and conclusion With the dominant semantic information in conclusion;

Latent Semantic information extraction modules 203, for there are Latent Semantic information when the known conditions and/or conclusion When, extract the Latent Semantic information in the known conditions and/or conclusion;

Merging module 204, for merging the dominant semantic information and institute that the dominant Semantic features extraction module is extracted The Latent Semantic information that Latent Semantic information extraction modules are extracted is stated, the semantic information of the stem is obtained.

Optionally, described device further includes judgment module, is used for:

According to the known conditions and/or conclusion, according to the obtained semantic information of the extracting mode of dominant semantic information In when having entity missing, then judging the known conditions and/or conclusion, there are Latent Semantic information;

The Latent Semantic information extraction modules 203 are connected with the judgment module, for judging institute when judgment module It states in known conditions and/or conclusion there are when Latent Semantic information, extracts the Latent Semantic in the known conditions and/or conclusion Information;

The merging module 204 is specifically used for merging the dominant language that the dominant Semantic features extraction module 202 is extracted The Latent Semantic information that adopted information and the Latent Semantic information extraction modules 203 are extracted obtains the semantic letter of the stem Breath.

Optionally, the dominant Semantic features extraction module 202 is specifically used for:

According to the correspondence table of keyword and relationship, the corresponding relationship of keyword in the known conditions and conclusion is obtained;

According to the positional relationship between the entity in the keyword and the known conditions and conclusion, in conjunction with the key The corresponding relationship of word extracts the dominant semantic information in the known conditions and conclusion.

Optionally, the Latent Semantic information extraction modules 203 are specifically used for:

When having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information, lacked according to existing Relationship in the semantic information of unfounded body obtains the corresponding keyword of the relationship;

According to the sentence before or after the keyword, the recessive language in the known conditions and/or conclusion is extracted Adopted information, and,

When extracting the known conditions and/or knot according to the sentence before or after the sentence where the keyword During Latent Semantic information in, if having found the supplement semantic information of the stem, the supplement is extracted Semantic information, and the supplement semantic information is denoted as Latent Semantic information.

In the embodiment of the present application, division module 201, dominant Semantic features extraction module 202, Latent Semantic information extraction mould Block 203, merging module 204 can be used for realizing step corresponding in above method embodiment.

It should be noted that the embodiment of the present application can also first obtain the entity in known conditions and/or conclusion, then obtain Relative in known conditions and/or conclusion, then according to the entity in obtained relative and known conditions and/or conclusion it Between relative position, extract the dominant semantic information in known conditions and/or conclusion.I.e. the embodiment of the present application is to from known conditions And/or relative is obtained in conclusion and is not especially limited with the sequencing for obtaining entity.

In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, is not described in detail or remembers in some embodiment The part of load, reference can be made to the related descriptions of other embodiments.

Those of ordinary skill in the art may be aware that function described in conjunction with the examples disclosed in the embodiments of the present disclosure Can module and method and step, can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions It is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Professional technique Personnel can specifically realize described function to each using distinct methods, but this realization is it is not considered that super The scope of the present invention out.

In embodiment provided by the present invention, it should be understood that disclosed device and method can pass through others Mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of above-mentioned module or unit, Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be with In conjunction with or be desirably integrated into another system, or some features can be ignored or not executed.Another point, it is shown or discussed Mutual coupling or direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING of device or unit or Communication connection can be electrical property, mechanical or other forms.

Above-mentioned unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme 's.

If above-mentioned integrated unit is realized in the form of SFU software functional unit and sells or use as independent product When, it can store in a computer readable storage medium.Based on this understanding, the present invention realizes above-described embodiment side All or part of the process in method can also instruct relevant hardware to complete, above-mentioned computer by computer program Program can be stored in a computer readable storage medium, and the computer program is when being executed by processor, it can be achieved that above-mentioned each The step of a embodiment of the method.Wherein, above-mentioned computer program includes computer program code, and above-mentioned computer program code can Think source code form, object identification code form, executable file or certain intermediate forms etc..Above-mentioned computer readable storage medium It may include: any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic that can carry above-mentioned computer program code Dish, CD, computer storage, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that above-mentioned It is appropriate that the content that computer readable storage medium includes can be carried out according to the requirement made laws in jurisdiction with patent practice Increase and decrease, such as do not include electric carrier wave according to legislation and patent practice, computer readable storage medium in certain jurisdictions Signal and telecommunication signal.

Above-described embodiment is merely illustrative of the technical solution of the present invention, rather than its limitations.Although the present invention has been described The preferred embodiment of embodiment, but one of ordinary skilled in the art once knows basic creative concept, then it can be to this A little embodiments make other change and modification.So the following claims are intended to be interpreted as including preferred embodiment and falls Enter all change and modification of range of embodiment of the invention.Obviously, those skilled in the art can to the embodiment of the present invention into Spirit and scope of the various modification and variations of row without departing from the embodiment of the present invention.If in this way, these of the embodiment of the present invention Modifications and variations belong within the scope of claim of the embodiment of the present invention and its equivalent technologies, then the embodiment of the present invention is also intended to It includes these modifications and variations.

Claims (10)

1. a kind of Semantic Information Processing method, which comprises the following steps:
According to obtained stem, the stem is divided into known conditions and conclusion two parts;
According to obtained known conditions and conclusion, the dominant semantic information in the known conditions and conclusion is extracted;
When, there are when Latent Semantic information, being extracted in the known conditions and/or conclusion in the known conditions and/or conclusion Latent Semantic information;
Merge the dominant semantic information and Latent Semantic information extracted, obtains the semantic information of the stem.
2. a kind of Semantic Information Processing method according to claim 1, which is characterized in that the known conditions and/or knot By there are the judgment steps of Latent Semantic information are as follows:
According to the known conditions and/or conclusion, according to having in the obtained semantic information of the extracting mode of dominant semantic information When entity lacks, then judging the known conditions and/or conclusion, there are Latent Semantic information.
3. a kind of Semantic Information Processing method according to claim 1, which is characterized in that the known item that the basis obtains Part and conclusion, the step of extracting the dominant semantic information in the known conditions and conclusion specifically:
According to the correspondence table of keyword and relationship, the corresponding relationship of keyword in the known conditions and conclusion is obtained;
According to the positional relationship between the entity in the keyword and the known conditions and conclusion, in conjunction with the keyword pair The relationship answered extracts the dominant semantic information in the known conditions and conclusion.
4. a kind of Semantic Information Processing method according to claim 1 or 2 or 3, which is characterized in that when the known conditions And/or conclusion extracts the specific step of the Latent Semantic information in the known conditions and/or conclusion there are when Latent Semantic information Suddenly are as follows:
It is real according to there is missing when having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information Relationship in the semantic information of body obtains the corresponding keyword of the relationship;
According to the sentence before or after the keyword, the Latent Semantic letter in the known conditions and/or conclusion is extracted Breath.
5. a kind of Semantic Information Processing method according to claim 1 or 3, which is characterized in that the basis obtains The step of knowing condition and conclusion, extracting the dominant semantic information in the known conditions and conclusion specifically:
According to the comma in the known conditions and conclusion, obtain in the known conditions and conclusion divided by the comma Single statement;
Dominant semantic information is extracted from the single statement respectively;
Merge the dominant semantic information extracted from each single statement, obtains aobvious in the known conditions and conclusion Property semantic information.
6. a kind of Semantic Information Processing method according to claim 4, which is characterized in that the merging is extracted dominant Semantic information and Latent Semantic information, before obtaining the semantic information of the stem further include:
When extracting in the known conditions and/or conclusion according to the sentence before or after the sentence where the keyword Latent Semantic information during, if having found the supplement semantic information of the stem, it is semantic to extract the supplement Information, and the supplement semantic information is denoted as Latent Semantic information.
7. a kind of Semantic Information Processing device characterized by comprising
Division module, for according to obtained stem, the stem to be divided into known conditions and conclusion two parts;
Dominant Semantic features extraction module, for extracting the known conditions and conclusion according to obtained known conditions and conclusion In dominant semantic information;
Latent Semantic information extraction modules, for extracting institute when the known conditions and/or conclusion are there are when Latent Semantic information State the Latent Semantic information in known conditions and/or conclusion;
Merging module, for merging the dominant semantic information and the recessive language that the dominant Semantic features extraction module is extracted The Latent Semantic information that adopted information extraction modules are extracted, obtains the semantic information of the stem.
8. a kind of Semantic Information Processing device according to claim 7, which is characterized in that described device further includes judging mould Block is used for:
According to the known conditions and/or conclusion, according to having in the obtained semantic information of the extracting mode of dominant semantic information When entity lacks, then judging the known conditions and/or conclusion, there are Latent Semantic information;
The Latent Semantic information extraction modules are connected with the judgment module, for judging the known item when judgment module There are when Latent Semantic information in part and/or conclusion, the Latent Semantic information in the known conditions and/or conclusion is extracted;
The merging module is specifically used for merging the dominant semantic information and institute that the dominant Semantic features extraction module is extracted The Latent Semantic information that Latent Semantic information extraction modules are extracted is stated, the semantic information of the stem is obtained.
9. a kind of Semantic Information Processing device according to claim 7 or 8, which is characterized in that the dominant semantic information Extraction module is specifically used for:
According to the correspondence table of keyword and relationship, the corresponding relationship of keyword in the known conditions and conclusion is obtained;
According to the positional relationship between the entity in the keyword and the known conditions and conclusion, in conjunction with the keyword pair The relationship answered extracts the dominant semantic information in the known conditions and conclusion.
10. a kind of Semantic Information Processing device according to claim 7 or 8, which is characterized in that the Latent Semantic information Extraction module is specifically used for:
It is real according to there is missing when having entity missing in the obtained semantic information of the extracting mode according to dominant semantic information Relationship in the semantic information of body obtains the corresponding keyword of the relationship;
According to the sentence before or after the keyword, the Latent Semantic letter in the known conditions and/or conclusion is extracted Breath, and,
When extracting in the known conditions and/or conclusion according to the sentence before or after the sentence where the keyword Latent Semantic information during, if having found the supplement semantic information of the stem, it is semantic to extract the supplement Information, and the supplement semantic information is denoted as Latent Semantic information.
CN201810081517.6A 2018-01-29 2018-01-29 A kind of Semantic Information Processing method and apparatus CN110096692A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810081517.6A CN110096692A (en) 2018-01-29 2018-01-29 A kind of Semantic Information Processing method and apparatus

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810081517.6A CN110096692A (en) 2018-01-29 2018-01-29 A kind of Semantic Information Processing method and apparatus

Publications (1)

Publication Number Publication Date
CN110096692A true CN110096692A (en) 2019-08-06

Family

ID=67442648

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810081517.6A CN110096692A (en) 2018-01-29 2018-01-29 A kind of Semantic Information Processing method and apparatus

Country Status (1)

Country Link
CN (1) CN110096692A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473224A (en) * 2013-09-30 2013-12-25 成都景弘智能科技有限公司 Exercise semantization method based on problem solving process
US20150046792A1 (en) * 2013-08-06 2015-02-12 Educational Testing Service System and Method for Rendering an Assessment Item
CN103250149B (en) * 2010-12-07 2015-11-25 Sk电信有限公司 For extracting semantic distance and according to the method for semantic distance to mathematics statement classification and the device for the method from mathematics statement
CN105302875A (en) * 2015-10-09 2016-02-03 电子科技大学 Interactive problem solving method, server and system
CN105488015A (en) * 2016-01-22 2016-04-13 吴以凡 Test question structuring method
CN107562918A (en) * 2017-09-12 2018-01-09 北京点易通科技有限公司 A kind of mathematical problem knowledge point discovery and batch label acquisition method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103250149B (en) * 2010-12-07 2015-11-25 Sk电信有限公司 For extracting semantic distance and according to the method for semantic distance to mathematics statement classification and the device for the method from mathematics statement
US20150046792A1 (en) * 2013-08-06 2015-02-12 Educational Testing Service System and Method for Rendering an Assessment Item
CN103473224A (en) * 2013-09-30 2013-12-25 成都景弘智能科技有限公司 Exercise semantization method based on problem solving process
CN105302875A (en) * 2015-10-09 2016-02-03 电子科技大学 Interactive problem solving method, server and system
CN105488015A (en) * 2016-01-22 2016-04-13 吴以凡 Test question structuring method
CN107562918A (en) * 2017-09-12 2018-01-09 北京点易通科技有限公司 A kind of mathematical problem knowledge point discovery and batch label acquisition method

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
李幸: "基于语义题库的智能教辅系统", 《中国优秀硕士学位论文全文数据库 社会科学Ⅱ辑》 *
李瞿男: "基于本体的语义题库的构建与知识挖掘的研究及应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Similar Documents

Publication Publication Date Title
Pelevina et al. Making sense of word embeddings
Antoniol et al. Design pattern recovery in object-oriented software
US9020805B2 (en) Context-based disambiguation of acronyms and abbreviations
US8359327B2 (en) Document processing method and system
Yue et al. A similarity measure based on species proportions
US20150154286A1 (en) Method for disambiguated features in unstructured text
US8209665B2 (en) Identification of topics in source code
Soibelman et al. Management and analysis of unstructured construction data types
CN101097573B (en) Automatically request-answering system and method
US9626622B2 (en) Training a question/answer system using answer keys based on forum content
CN103229168B (en) The method and system that evidence spreads between multiple candidate answers during question and answer
Pennacchiotti et al. Automatic induction of FrameNet lexical units
JP5924666B2 (en) Predicate template collection device, specific phrase pair collection device, and computer program therefor
US9740771B2 (en) Information handling system and computer program product for deducing entity relationships across corpora using cluster based dictionary vocabulary lexicon
CN103221952A (en) Lexical answer type confidence estimation and application
CN104750798B (en) Recommendation method and device for application program
CN105653706A (en) Multilayer quotation recommendation method based on literature content mapping knowledge domain
CN102591914B (en) Identifying key phrases within documents
KR20140131327A (en) Social media data analysis system and method
US20030208285A1 (en) Method for comparing solid models
EP2991003A2 (en) Method and apparatus for classification
CN105183923B (en) New word discovery method and device
CN102609406B (en) Learning device, judgment means, learning method and determination methods
CN102722709B (en) Method and device for identifying garbage pictures
Bing et al. A fuzzy logic approach for opinion mining on large scale twitter data

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