CN105824798A - Examination question de-duplicating method of examination question base based on examination question key word likeness - Google Patents
Examination question de-duplicating method of examination question base based on examination question key word likeness Download PDFInfo
- Publication number
- CN105824798A CN105824798A CN201610117476.2A CN201610117476A CN105824798A CN 105824798 A CN105824798 A CN 105824798A CN 201610117476 A CN201610117476 A CN 201610117476A CN 105824798 A CN105824798 A CN 105824798A
- Authority
- CN
- China
- Prior art keywords
- examination question
- examination
- question
- keyword
- relational database
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to an examination question de-duplicating method of an examination question base based on examination question key word likeness. The examination question de-duplicating method comprises the following steps: firstly, performing Chinese word segmentation on examination questions so as to obtain word segmentation knots; judging whether the word segmentation knots are key words or not, if yes, adding the word segmentation knots to a relational database of the examination questions and the key words; then calculating the likeness of any two examination questions to be detected in the relational database of the examination questions and the key words by using a scalar product; secondly, judging whether the two examination questions to be detected are non-like examination questions and adding the like examination questions into a duplication examination question relational database; searching a duplication examination question list from the duplication examination question relational database according to the likeness condition; finally, confirming the duplication examination questions by observing a duplication examination question list by an administrator so as to judge whether the examination questions are duplicated manually. According to the examination question de-duplicating method disclosed by the invention, Chinese word segmentation is performed on question stems, examination question candidate items and examination question answers of the examination questions, segmented words after word segmentation is performed are analyzed, and the examination questions are deeply analyzed, so that de-duplication is accurate. The method disclosed by the invention can be widely used in the field of de-duplication of the examination questions.
Description
Technical field
The present invention relates to a kind of examination question De-weight method, especially with regard to the examination question De-weight method in a kind of test item bank based on examination question keyword similarity.
Background technology
Along with carrying out of all kinds of works about test over the years, inside test item bank, the exercise question of accumulation also gets more and more, and has gradually formed the test item bank of magnanimity.Owing to the examination question in the test item bank of certain a branch of instruction in school is in different periods, formed by the expert of different majors and varying level writing of making joint efforts, result in all kinds of forms of appearance different, such as multiple-choice question, filling topic, True-False conciliate answer etc., and there is different difficulties, but the repetition examination question that implication is similar or identical, although the form of expression repeating examination question may be more, but can be attributed to following two classes:
(1) examination question that examination question word content is identical or word content is the most close and answer is identical；
(2) character express of examination question is different or topic type is different but examination knowledge is identical；
Repetition item analysis for existing Test System judges to repeat examination question only by analysis stem word is the most identical, the analysis ability repeating examination question from stem character analysis is very limited, hinders problems such as examination point analysis, test papers and the building-up of question banks.It addition, only by analyzing, stem word is the most identical to be judged to repeat examination question comprehensively, repeats examination question discrimination the highest, and precision is inadequate.
Summary of the invention
For the problems referred to above, it is an object of the invention to provide the examination question De-weight method in a kind of test item bank based on examination question keyword similarity, the highest to improve repetition examination question discrimination, and the problem that precision is inadequate.
For achieving the above object, the present invention takes techniques below scheme: the examination question De-weight method in a kind of test item bank based on examination question keyword similarity, it comprises the following steps: 1) use maximum forward participle matching algorithm that the examination question in test item bank is carried out Chinese word segmentation, Chinese word segmentation includes the stem of examination question, examination question candidate item and script in test item bank are carried out Chinese word segmentation, and the participle obtained is referred to as participle knot；Judge whether participle knot is the key word in examination question keywords database, if the key word in examination question keywords database, then it is added into the relational database of examination question and keyword, and the relational database of examination question and keyword includes the order that the frequency of occurrences of keyword, keyword weights and keyword occur；Wherein, examination question key word library presets examination question keyword；2) similarity between any two examination question to be detected in the relational database of inner product calculating examination question and keyword is used；3) by within product representation similarity with repeat compared with examination question threshold value, if repetition examination question threshold value the most set in advance, then perform step 4)；If more than repetition examination question threshold value set in advance, then perform step 5)；4) two examination questions to be detected are non-similar examination question, do not process；5) two examination questions to be detected are similar examination question, and similar examination question adds repetition examination question relational database；6) according to similarity condition, from repetition examination question relational database, the repetition examination question list meeting condition is found out；7) management personnel carry out repeating examination question confirmation by valuing retrial topic list, artificially judge whether examination question repeats.
Described step 3) in, it is judged that within whether examination question repeats, the repetition examination question threshold value of product representation is 0.80.
Due to the fact that and take above technical scheme, it has the advantage that the present invention carries out Chinese word segmentation initially with maximum forward participle matching algorithm to the examination question in test item bank, and the participle obtained is referred to as participle knot；Judge whether participle knot is the key word in examination question keywords database, if the key word in examination question keywords database, is then added into the relational database of examination question and keyword；Then, the similarity between any two examination question to be detected in the relational database of inner product calculating examination question and keyword is used；Secondly, by within the similarity of product representation compared with repeating examination question threshold value, it is judged that whether two examination questions to be detected are non-similar examination question, and similar examination question adds repetition examination question relational database；Again, according to similarity condition, from repetition examination question relational database, the repetition examination question list meeting condition is found out；Finally, management personnel carry out repeating examination question confirmation by valuing retrial topic list, artificially judge whether examination question repeats.Processing step by above, the present invention not only stem to examination question carries out Chinese Word Segmentation, also examination question candidate item and script is carried out Chinese word segmentation, analyzes for the participle after cutting word comprehensively, thus analyse in depth examination question.It addition, use the mode of manual confirmation can improve the accuracy rate heavily inscribing judgement, improve and remove the precision heavily inscribed.Therefore, present invention can be widely used to examination question duplicate removal field.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, the accompanying drawing used required in embodiment or description of the prior art will be briefly described below, apparently, accompanying drawing in describing below is only some embodiments of the present invention, for those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the flow chart of the present invention.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only a part of embodiment of the present invention rather than whole embodiments.Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art are obtained under not making creative work premise, broadly fall into the scope of protection of the invention.
Embodiment
As it is shown in figure 1, the examination question De-weight method in a kind of test item bank based on examination question keyword similarity of the present invention, it comprises the following steps:
1) using maximum forward participle matching algorithm that the examination question in test item bank is carried out Chinese word segmentation, Chinese word segmentation includes the stem of examination question, examination question candidate item and script in test item bank are carried out Chinese word segmentation, and the participle obtained is referred to as participle knot；Judge whether participle knot is the key word in examination question keywords database, if the key word in examination question keywords database, then it is added into the relational database of examination question and keyword, and the relational database of examination question and keyword includes the order that the frequency of occurrences of keyword, keyword weights and keyword occur；Wherein, examination question key word library presets examination question keyword, such as T1, T2..., Tm。
It should be noted that the algorithm that maximum forward participle matching algorithm is known to the skilled person, therefore no longer describe in detail.
A, employing maximum forward participle matching algorithm carry out Chinese word segmentation to the examination question in test item bank, and Chinese word segmentation includes the stem of examination question, examination question candidate item and script in test item bank are carried out Chinese word segmentation, and the participle obtained is referred to as participle knot；
The embodiment of the present invention uses maximum forward participle matching algorithm that examination question is carried out Chinese word segmentation, maximum forward participle matching algorithm is from left to right to be mated with vocabulary by the several continuation characters treating in participle examination question, if the character obtained matches with the word in vocabulary, then it is syncopated as a word；Otherwise, do not process.If wanting to accomplish maximum match, not the most that once coupling can be carried out cutting, being illustrated by example below:
nullTreat participle examination question: content []={ " directly "，" line "，" bar "，" tower "，" "，" hang down "，" straight "，" shelves "，" away from "，" more "，" big "，" absolutely "，" edge "，" son "，" string "，" institute "，" by "，" "，" lotus "，" weight "，" just "，" more "，" big " }，I.e. content [1] is " directly "，Content [2] is " line "，Content [3] is " bar "，Content [4] is " tower "，Content [5] be " "，Content [6] is " hanging down "，Content [7] is " directly "，Content [8] is " shelves "，Content [9] be " away from "，Content [10] is " getting over "，Content [11] is " greatly "，Content [12] is " absolutely "，Content [13] is " edge "，Content [14] is " sub "，Content [15] is " institute "，Content [16] is " being subject to "，Content [17] be " "，Content [18] is " lotus "，Content [19] is " weight "，Content [20] is " just "，Content [21] is " getting over "，Content [22] is " greatly ".
Vocabulary: vocabulary dict []={ " straight line ", " shaft tower ", " straight line pole ", " insulator " }, wherein, dict [1] is " straight line ", and dict [2] is " shaft tower ", and dict [3] is " straight line pole ", and dict [4] is " insulator ".
As follows for the maximum forward participle matching algorithm solution procedure treating participle examination question:
1., from the beginning of content [1], when scanning content [2] when, find that " straight line " suffers at vocabulary dict [], therefore can not cut out, because not knowing whether obtained word is longer word, i.e. maximum match, it is therefore desirable to continue to scan on；
2. content [3] is continued to scan on, find that " straight line pole " is not the word in vocabulary dict [], but can't determine that " straight line " that above find has been maximum word, because " straight line pole " is the prefix of dict [3], it is therefore desirable to continue to scan on；
3. continue to scan on content [4], find that " straight line pole " is the word in vocabulary dict [], but still can not cut out, because not knowing whether obtained word is longer word, i.e. maximum match, it is therefore desirable to continue to scan on；
4. continue to scan on content [5], find that " straight line pole " is not the word in vocabulary, be not the prefix of word.Therefore the most maximum word can be syncopated as " straight line pole ".
Understanding for the maximum forward participle matching algorithm solution procedure treating participle examination question, the word that maximum match goes out must assure that next scanning is not that the prefix of the word in vocabulary or word just can terminate.
B, judge participle knot whether be the key word in examination question keywords database, if the key word in examination question keywords database, then it is added into the relational database of examination question and keyword, and the relational database of examination question and keyword includes the order that the frequency of occurrences of keyword, keyword weights and keyword occur；
Trie tree construction is used to be stored with examination question by keyword, this kind of mode is used to store, making the time complexity searching each word is O (word.length), and can judge whether that the match is successful or has matched the prefix of character string very easily.Storage organization is:
The most each node is a Chinese character in word；
2. the pointer in node has pointed to this Chinese character next Chinese character in some word.These pointers leave in the hash structure with Chinese character as key；
3. the Chinese character during " # " in node represents current node is the last character of the word formed to this Chinese character node from root node.
2) similarity between any two examination question to be detected in the relational database of inner product calculating examination question and keyword is used；
In tradition vector space model, it is elementary composition vector that the examination question to be detected that examination question to be detected compares with it is all expressed as with examination question key word, each examination question keyword root is according to word frequency TF and inverse text frequency IDF (TF-IDF, Termfrequency-inversedocumentfrequency, word frequency-inverted file frequency) it is endowed certain weights, then by cosine angle between vector element or the similarity calculated between examination question to be detected of other parameter, the similarity asking the method for co sinus vector included angle to obtain between examination question to be detected in Euclidean space is used.
In examination question vector space model, per pass examination question is by separate key word T1, T2..., TmConstitute, make D=(D1, D2..., Dn) it is the set of n the examination question that m indexing key words is constituted, wherein Dj=(d1j, d2j..., dmj)TIt is examination question vector, dijRepresent that key word i occurs the frequency weight in examination question, and 1≤i≤m, query vector Q are expressed as Q=(q1, q2..., qm)T, qiRepresent that frequency weight in queries occurs in key word i, this defines a m and tie up key words content vector space, i.e. examination question keyword vector space.
To examination question Similarity Measure to be detected, we are calculated by inner product formula, if DiWith DjIt is the examination question that in the set D of examination question, any two differs, and Di=(d1i, d2i..., dmi)T, Dj=(d1j, d2j..., dmj)T, then DiWith DjBetween similarity inner product be expressed as follows:
Wherein, dkiThe frequency weight occurred in examination question to be detected for key word k, and 1≤k≤m.
3) by within product representation similarity with repeat compared with examination question threshold value, if repetition examination question threshold value the most set in advance, then perform step 4)；If more than repetition examination question threshold value set in advance, then perform step 5)；
Above-mentioned included angle cosine is used for measuring the size of angle between two groups of vectors, also known as phase and coefficient, and the geometric meaning of included angle cosine is by N number of elementary composition N-dimensional space, characterizes the cosine value of angle between two vectors.Typically needing before use each element in vector is carried out nondimensionalization process, making each element is just all, and at this moment the span of included angle cosine is [0,1], and value shows that the most greatly two vector angles are the least, both closer to, when value is 1, two vector identical.
As Sim (Di, DjDuring)≤0.80, then perform step 4)；
As Sim (Di, Dj) ＞ 0.80 time, then perform step 5)；
4) two examination questions to be detected are non-similar examination question, do not process；
5) two examination questions to be detected are similar examination question, and similar examination question adds repetition examination question relational database；
6) according to similarity condition, from repetition examination question relational database, the repetition examination question list meeting condition is found out；
Wherein, similarity condition is condition well known to those skilled in the art, therefore no longer describes in detail.
7) management personnel carry out repeating examination question confirmation by valuing retrial topic list, artificially judge whether examination question repeats, and use the mode of manual confirmation can improve the accuracy rate heavily inscribing judgement, improve and remove the precision heavily inscribed.
The various embodiments described above are merely to illustrate the present invention; the structure of the most each parts, connected mode and processing technology etc. all can be varied from; every equivalents carried out on the basis of technical solution of the present invention and improvement, the most should not get rid of outside protection scope of the present invention.
Claims (2)
1. the examination question De-weight method in test item bank based on examination question keyword similarity, it comprises the following steps:
1) using maximum forward participle matching algorithm that the examination question in test item bank is carried out Chinese word segmentation, Chinese word segmentation includes the stem of examination question, examination question candidate item and script in test item bank are carried out Chinese word segmentation, and the participle obtained is referred to as participle knot；Judge whether participle knot is the key word in examination question keywords database, if the key word in examination question keywords database, then it is added into the relational database of examination question and keyword, and the relational database of examination question and keyword includes the order that the frequency of occurrences of keyword, keyword weights and keyword occur；
Wherein, examination question key word library presets examination question keyword；
2) similarity between any two examination question to be detected in the relational database of inner product calculating examination question and keyword is used；
3) by within product representation similarity with repeat compared with examination question threshold value, if repetition examination question threshold value the most set in advance, then perform step 4)；If more than repetition examination question threshold value set in advance, then perform step 5)；
4) two examination questions to be detected are non-similar examination question, do not process；
5) two examination questions to be detected are similar examination question, and similar examination question adds repetition examination question relational database；
6) according to similarity condition, from repetition examination question relational database, the repetition examination question list meeting condition is found out；
7) management personnel carry out repeating examination question confirmation by valuing retrial topic list, artificially judge whether examination question repeats.
Examination question De-weight method in test item bank based on examination question keyword similarity the most according to claim 1, it is characterised in that: described step 3) in, it is judged that within whether examination question repeats, the repetition examination question threshold value of product representation is 0.80.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610117476.2A CN105824798A (en) | 2016-03-03 | 2016-03-03 | Examination question de-duplicating method of examination question base based on examination question key word likeness |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610117476.2A CN105824798A (en) | 2016-03-03 | 2016-03-03 | Examination question de-duplicating method of examination question base based on examination question key word likeness |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105824798A true CN105824798A (en) | 2016-08-03 |
Family
ID=56988063
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610117476.2A Pending CN105824798A (en) | 2016-03-03 | 2016-03-03 | Examination question de-duplicating method of examination question base based on examination question key word likeness |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105824798A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106326417A (en) * | 2016-08-24 | 2017-01-11 | 冯玉国 | Test question data processing method and system |
| CN106815372A (en) * | 2017-02-06 | 2017-06-09 | 广东小天才科技有限公司 | Test question duplication removing method and device based on physical examination question bank and user equipment |
| CN108664630A (en) * | 2018-05-14 | 2018-10-16 | 广西英腾教育科技股份有限公司 | Examination question De-weight method and device |
| CN108763476A (en) * | 2018-05-29 | 2018-11-06 | 深圳市三宝创新智能有限公司 | A kind of question and answer Data clean system based on part of speech weight calculation |
| CN109086313A (en) * | 2018-06-27 | 2018-12-25 | 马赫 | One kind carrying out examination question based on inverse text similarity and orders rearrangement processed |
| CN109241395A (en) * | 2018-06-27 | 2019-01-18 | 广州市南方人力资源评价中心有限公司 | A kind of examination question network re-scheduling retrieval method based on keyword resolution |
| CN110390019A (en) * | 2019-07-26 | 2019-10-29 | 江苏曲速教育科技有限公司 | A kind of clustering method of examination question, De-weight method and system |
| CN111027321A (en) * | 2019-11-30 | 2020-04-17 | 南京森林警察学院 | Police affair related intelligent question-making system |
| CN111241239A (en) * | 2020-01-07 | 2020-06-05 | 科大讯飞股份有限公司 | Method for detecting repeated questions, related device and readable storage medium |
| CN111459970A (en) * | 2020-03-31 | 2020-07-28 | 交通银行股份有限公司 | Method for checking uniqueness of object information |
| CN111625468A (en) * | 2020-06-05 | 2020-09-04 | 中国银行股份有限公司 | Test case duplicate removal method and device |
| CN112069295A (en) * | 2020-09-18 | 2020-12-11 | 科大讯飞股份有限公司 | Similar question recommendation method and device, electronic equipment and storage medium |
| CN112115232A (en) * | 2020-09-24 | 2020-12-22 | 腾讯科技（深圳）有限公司 | Data error correction method and device and server |
| CN112216168A (en) * | 2020-10-21 | 2021-01-12 | 李帮军 | Intelligent question type conversion system and method based on choice question editor |
| CN113076734A (en) * | 2021-04-15 | 2021-07-06 | 云南电网有限责任公司电力科学研究院 | Similarity detection method and device for project texts |
| CN114048354A (en) * | 2022-01-10 | 2022-02-15 | 广州启辰电子科技有限公司 | Test question retrieval method, device and medium based on multi-element characterization and metric learning |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102629272A (en) * | 2012-03-14 | 2012-08-08 | 北京邮电大学 | Clustering based optimization method for examination system database |
| CN103136302A (en) * | 2011-12-05 | 2013-06-05 | 北大方正集团有限公司 | Method and device of test question repeat output |
-
2016
- 2016-03-03 CN CN201610117476.2A patent/CN105824798A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103136302A (en) * | 2011-12-05 | 2013-06-05 | 北大方正集团有限公司 | Method and device of test question repeat output |
| CN102629272A (en) * | 2012-03-14 | 2012-08-08 | 北京邮电大学 | Clustering based optimization method for examination system database |
Non-Patent Citations (2)
| Title |
|---|
| 汤世平 等: "基于多示例学习的题库重复性检测研究", 《北京理工大学学报》 * |
| 程维刚 等: "基于关键词匹配技术的相似试题检测方法研究", 《北华航天工业学院学报》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106326417A (en) * | 2016-08-24 | 2017-01-11 | 冯玉国 | Test question data processing method and system |
| CN106815372A (en) * | 2017-02-06 | 2017-06-09 | 广东小天才科技有限公司 | Test question duplication removing method and device based on physical examination question bank and user equipment |
| CN108664630A (en) * | 2018-05-14 | 2018-10-16 | 广西英腾教育科技股份有限公司 | Examination question De-weight method and device |
| CN108763476A (en) * | 2018-05-29 | 2018-11-06 | 深圳市三宝创新智能有限公司 | A kind of question and answer Data clean system based on part of speech weight calculation |
| CN109241395B (en) * | 2018-06-27 | 2021-08-03 | 广州市南方人力资源评价中心有限公司 | Keyword analysis-based test question network duplicate elimination and retrieval method |
| CN109086313A (en) * | 2018-06-27 | 2018-12-25 | 马赫 | One kind carrying out examination question based on inverse text similarity and orders rearrangement processed |
| CN109241395A (en) * | 2018-06-27 | 2019-01-18 | 广州市南方人力资源评价中心有限公司 | A kind of examination question network re-scheduling retrieval method based on keyword resolution |
| CN110390019A (en) * | 2019-07-26 | 2019-10-29 | 江苏曲速教育科技有限公司 | A kind of clustering method of examination question, De-weight method and system |
| CN111027321A (en) * | 2019-11-30 | 2020-04-17 | 南京森林警察学院 | Police affair related intelligent question-making system |
| CN111241239A (en) * | 2020-01-07 | 2020-06-05 | 科大讯飞股份有限公司 | Method for detecting repeated questions, related device and readable storage medium |
| CN111241239B (en) * | 2020-01-07 | 2022-12-02 | 科大讯飞股份有限公司 | Method for detecting repeated questions, related device and readable storage medium |
| CN111459970A (en) * | 2020-03-31 | 2020-07-28 | 交通银行股份有限公司 | Method for checking uniqueness of object information |
| CN111625468A (en) * | 2020-06-05 | 2020-09-04 | 中国银行股份有限公司 | Test case duplicate removal method and device |
| CN111625468B (en) * | 2020-06-05 | 2024-04-16 | 中国银行股份有限公司 | Test case duplicate removal method and device |
| CN112069295A (en) * | 2020-09-18 | 2020-12-11 | 科大讯飞股份有限公司 | Similar question recommendation method and device, electronic equipment and storage medium |
| CN112115232A (en) * | 2020-09-24 | 2020-12-22 | 腾讯科技（深圳）有限公司 | Data error correction method and device and server |
| CN112216168A (en) * | 2020-10-21 | 2021-01-12 | 李帮军 | Intelligent question type conversion system and method based on choice question editor |
| CN113076734A (en) * | 2021-04-15 | 2021-07-06 | 云南电网有限责任公司电力科学研究院 | Similarity detection method and device for project texts |
| CN114048354A (en) * | 2022-01-10 | 2022-02-15 | 广州启辰电子科技有限公司 | Test question retrieval method, device and medium based on multi-element characterization and metric learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105824798A (en) | Examination question de-duplicating method of examination question base based on examination question key word likeness | |
| CN109635461B (en) | Method and system for automatically identifying surrounding rock grade by using while-drilling parameters | |
| CN108132927B (en) | Keyword extraction method for combining graph structure and node association | |
| CN109408743A (en) | Text link embedding grammar | |
| CN109271511B (en) | Automatic problem solving method based on complex reasoning network | |
| CN110782892B (en) | Voice text error correction method | |
| CN108268600A (en) | Unstructured Data Management and device based on AI | |
| CN106649662A (en) | Construction method of domain dictionary | |
| CN111368096A (en) | Knowledge graph-based information analysis method, device, equipment and storage medium | |
| CN110188174A (en) | A kind of professional domain FAQ intelligent answer method excavated based on specialized vocabulary | |
| CN114048327A (en) | Automatic subjective question scoring method and system based on knowledge graph | |
| CN115599902A (en) | Oil-gas encyclopedia question-answering method and system based on knowledge graph | |
| CN112612909B (en) | Intelligent test paper quality evaluation method based on knowledge graph | |
| CN112362292B (en) | Method for anomaly detection of wind tunnel test data | |
| CN112749530B (en) | Text encoding method, apparatus, device and computer readable storage medium | |
| Klein et al. | Algorithmic programming language identification | |
| Soni et al. | Emotion based social media text classification using optimized improved ID3 classifier | |
| Lawrie et al. | On the value of bug reports for retrieval-based bug localization | |
| CN116090449B (en) | Entity relation extraction method and system for quality problem analysis report | |
| CN111597400A (en) | Computer retrieval system and method based on way-finding algorithm | |
| CN115688789B (en) | Entity relation extraction model training method and system based on dynamic labels | |
| Chien et al. | A hybrid approach for automatic schema matching | |
| CN114238595A (en) | Metallurgical knowledge question-answering method and system based on knowledge graph | |
| CN111274404A (en) | Small sample entity multi-field classification method based on man-machine cooperation | |
| CN113869428A (en) | Intrusion recognition model generation method, intrusion recognition model and intrusion detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160803 |
|
| WD01 | Invention patent application deemed withdrawn after publication |