CN111008529A - Chinese relation extraction method based on neural network - Google Patents
Chinese relation extraction method based on neural network Download PDFInfo
- Publication number
- CN111008529A CN111008529A CN201910669521.9A CN201910669521A CN111008529A CN 111008529 A CN111008529 A CN 111008529A CN 201910669521 A CN201910669521 A CN 201910669521A CN 111008529 A CN111008529 A CN 111008529A
- Authority
- CN
- China
- Prior art keywords
- model
- neural network
- information
- chinese
- relation extraction
- 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.)
- Granted
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 16
- 238000000034 method Methods 0.000 claims abstract description 42
- 230000007246 mechanism Effects 0.000 claims abstract description 6
- 230000015654 memory Effects 0.000 claims abstract description 5
- 235000019580 granularity Nutrition 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000007774 longterm Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 13
- 238000013527 convolutional neural network Methods 0.000 abstract description 11
- 238000002474 experimental method Methods 0.000 abstract description 7
- 230000002457 bidirectional effect Effects 0.000 abstract description 4
- 230000001419 dependent effect Effects 0.000 abstract description 3
- 239000010410 layer Substances 0.000 description 22
- 230000008569 process Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 239000013598 vector Substances 0.000 description 5
- 238000003062 neural network model Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000004880 explosion Methods 0.000 description 2
- 230000007787 long-term memory Effects 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 239000002356 single layer Substances 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a Chinese relation extraction method based on a neural network, which can effectively acquire the structural information and semantic information of sentences. In the relation extraction task, a single long-short term memory model can only learn the features of a certain dimensionality, and the convolutional neural network can learn the features of different dimensionalities by utilizing a plurality of convolutional kernels. Based on the two characteristics, the invention provides a multilayer bidirectional long-short term memory-attention model, and the method can automatically extract abstract features with different dimensions and dependent information from the original input by setting hidden layers with different sizes for the long-short term memory model, and captures global information by using an attention mechanism. Experiments show that compared with a multi-core convolutional neural network and a single long-short term memory-attention model, the method disclosed by the invention can obviously improve the Chinese relation extraction effect, and obtains 71.61% of F value on an ACE RDC2005 Chinese data set, so that a good effect is obtained, and the effectiveness of the method is proved.
Description
Technical Field
The invention relates to the field of information extraction, in particular to a Chinese relation extraction method based on a neural network. Belongs to the technical field of natural language processing and machine learning.
Background
With the development of artificial intelligence and technical explosion in the field of information extraction, entity relationship extraction has received attention from more and more scholars as an important research topic in the field of information extraction. The main purpose is to extract semantic relations between labeled entity pairs in sentences, namely to determine relation categories between entity pairs in unstructured texts on the basis of entity recognition and form structured data for storage and retrieval. The result of the entity relation extraction can be used for constructing a knowledge map or an ontology knowledge base, and data support can be provided for the construction of an automatic question-answering system. Besides, the entity relationship extraction has important research significance in the aspects of semantic network annotation, chapter understanding and machine translation.
Early relation extraction was mainly based on grammatical rules, by analyzing grammatical structures in sentences as the basis for relation generation. Although the method achieves good results, the recall rate is difficult to promote due to strict rules, professional grammatical knowledge and literature foundation are needed, and the applicability is not high. With the continuous development of the technology, the relation extraction method is divided into a supervised method, a semi-supervised method and an unsupervised method. Based on the content related to the invention, a supervised relationship extraction method is mainly researched. Supervised relational extraction can be mostly regarded as a classification problem, and there are two main methods in summary: a shallow structure model and a deep learning model.
The shallow structure generally has only one layer or no hidden layer nodes, such as a support vector machine, maximum entropy and the like. Shallow structures in relational extraction often use methods of feature engineering or kernel functions. The traditional method based on feature engineering mainly relies on a skillful design of feature set output through a language processing process. Most of these methods rely on either a large number of manually designed features or on carefully designed kernel functions. Despite the assistance of many excellent NLP tools, there is still a risk of performance degradation due to errors such as word segmentation inaccuracy and syntax parsing error. More importantly, the low portability of these carefully designed features or kernel functions greatly affects their scalability.
In recent years, a great deal of research has been made on relationship extraction based on deep learning. The method for extracting the relationships is based on models such as CNN and RNN, and achieves excellent effects. Many neural network-based methods show the advantages of neural networks over traditional shallow structures, but most of these results are achieved on a distributed balanced english dataset and use many external features as aids. The Chinese grammar structure is complex, and the language fuzzy phenomenon is more serious.
Disclosure of Invention
The invention provides a Chinese relation extraction method based on a neural network. The method can automatically extract abstract features with different dimensionalities and dependent information from original input by arranging hidden layers with different sizes for the long-term and short-term memory model, and captures global information by using an attention mechanism. Experiments show that compared with a multi-core convolutional neural network and a single long-short term memory-attention model, the method can obviously improve the Chinese relation extraction effect, and obtains a better result on the ACE RDC2005 Chinese data set, so that the effectiveness of the method is proved. The model frame is shown in figure 1.
The technical scheme of the invention is as follows: a Chinese relation extraction method based on a neural network comprises the following steps: constructing a BilSTMA unit, and extracting deep semantic information and global dependency information of a sentence; step two, constructing a Multi-BilSTMA model, and acquiring semantic information with dependency relationship of different granularities; and step three, verifying the validity of the method by using the real data.
The step 1 fully utilizes the advantages of a bidirectional long-term and short-term memory model (BilSTM) in the aspect of processing long-term dependence problems and the characteristic that an Attention mechanism (Attention) can capture global dependence information, and constructs a BilSTM unit (BilSTM-Attention) to extract deep semantic information and dependence information of sentences.
Step 2, setting hidden layers with different sizes in the BilSTA units, combining the BilSTA units with different sizes, and constructing a Multi-BilSTA model, wherein the model can obtain semantic information with dependency relationship with different granularities.
In the step 3, in order to verify the effectiveness of the method, an ACE RDC200 Chinese data set is used for verifying the identification effect of the method, so that the effectiveness of the method is verified.
Advantageous effects
The invention has the beneficial effects that:
in the invention, the key point is that the characteristics that the Multi-core CNN can learn different granularity characteristics are used for reference, a Multi-BilSTMA model is constructed by setting different sizes of BilSTMs by using a BilSTM and an Attention mechanism, and experiments prove that the method has excellent effect on an ACE RDC2005 Chinese data set.
The invention provides a Chinese relation extraction method based on a Multi-BilSTM-Attention neural network model. Experiments prove that the method shows higher performance on an ACE data set, and the effectiveness of the method is proved. The method provided by the invention effectively utilizes the characteristic that different granularity characteristics can be learned in the multi-core CNN neural network, and combines the characteristic with the BilSTM, thereby fully playing the characteristic of automatic characteristic extraction of the neural network model. A plurality of hidden layers with different sizes are arranged in a bidirectional BILSTM channel, so that feature sparsity can be prevented to a certain extent, semantic information of characters can be effectively acquired and utilized, and abstract features with different dimensions can be automatically acquired. On the basis, an Attention mechanism is added, local features and global features of sentences are utilized, and the weight is adjusted through the features, so that the noise is reduced, and the accuracy is improved.
The method provided by the invention combines the characteristic that a single long-short term memory model can only learn a certain specific dimensionality with the characteristic that a plurality of convolution kernels in a convolution neural network can learn different dimensionalities, provides a Multi-BilSTM-Attention model, obtains excellent performance in the aspect of Chinese relation extraction, and obtains good use effect.
Drawings
FIG. 1 is a neural network model of Multi-BilSTM-Attention according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings.
For a sentence with two entities, the task of relationship extraction is to extract candidate relationships between the two entities. The bidirectional long-short term memory neural network (BilSTM) model belongs to a variant of a Recurrent Neural Network (RNN), can effectively process long-distance information and avoid gradient explosion, and is combined to use due to the fact that the BilSTM and the Attention have good complementarity. However, a single, fixed BilSTM can only learn information of a particular dimension, and thus a Multi-BilSTM model is constructed by setting different BilSTMs. The model can learn features with dependent information in multiple dimensions.
First, the input layer of the model consists of word vectors mapped to the lookup table obtained by initialization. If the sentence length is L, the sentence mapped into the vector can be represented as: x ═ X1,x2,···,xL]Wherein x isi∈RDIs the ith word wiD is the dimension of the vector. If the dictionary size is V, the Embedding layer can be expressed as X ∈ RV×H. This process can be expressed as: x ═ embedding(s).
Next, the Multi-BiLSTMA layer in the present invention is composed of three BiLSTMA units. Wherein each BilSTMA unit consists of one layer of BilSTM and one layer of Attention. As shown in fig. 1(b), the BiLSTMA receives the data of the Embedding layer, uses a forward LSTM and a backward LSTM, to form a BiLSTM layer, which is used to extract features of deeper layers of the Embedding. This process is summarized as follows:
representing an element-by-element addition. Information on each time step in the Attention layer is laminated and combined with the BilsTM layer, and a signal with large influence on the extraction result is obtained through calculationAnd (4) information. This process can be summarized as: a ═ extension (h).
The next step is the fully connected layer of the model. After the outputs of the three BilSTMA units are spliced together, the information learned by the model is classified through a layer of full connection (Dense) layer, wherein the size of the hidden layer is the number of the relational types, namely 7. This process is summarized as follows: d ═ sense (a).
And finally, in order to obtain a better experimental effect, performing normalization processing on the output result of the full connection layer by using a softmax layer to obtain a final classification result. In general, this process can be summarized as: y ═ softmax (d).
The effectiveness of the method is verified by adopting real data, the selected data is an ACE RDC2005 standard Chinese data set, and the data is firstly preprocessed.
The invention adopts a publicly issued ACE RDC2005 Chinese data set to perform relationship extraction. After screening out irregular documents, the experiment totally uses 628 documents. This data set contains 6 entity relationship types (collectively, the positive examples), which are: "PART-WHOLE", "PHYS", "ORG-AFF", "GEN-AFF", "PER-SOC", and "ART". Since the relationships in the data set are directional, for example: if the entity pair (A, B) has an "ART" relationship in the dataset, but there is no relationship type labeled by any dataset between the entity pair (B, A), all such cases are collectively called negative examples, and the relationship type is labeled as "Other". Since the relationship extraction is mainly performed at the sentence level, ". ","! ","? ",": "these 5 Chinese punctuation marks cut the text in the dataset into sentences. Sentences without entity pairs are discarded, and sentences repeated between positive examples and negative examples are removed (because the same sentence cannot be both positive examples and negative examples), so that 1010056 sentences are obtained in total, wherein the 1010056 sentences comprise 9244 positive example sentences and 91812 negative example sentences. ACE RDC2005 chinese dataset is a dataset with an unbalanced distribution, each relationship type is not evenly distributed, especially with negative cases up to 90.85% in percentage. In order to more closely approach the real situation and reduce the influence caused by a large amount of negative example data, only the positive example results are evaluated in the evaluation.
Secondly, on the aspect of word vector processing, a method of randomly initializing the Lookup Table is adopted, the Lookup Table can be continuously adjusted in the training process, and the dimension of the word vector is set to be 100 dimensions. Because the neural network needs fixed input, the average length of sentences corresponding to each relationship type is analyzed. In order to balance the extraction effect and the training cost, a sentence length equal to 50 is selected as the maximum input length, sentences with sentence lengths lower than 50 are filled with '0' to 50, and the cuts with sentence lengths higher than 50 are cut to 50. An AdaDelta function is selected as an optimization function, and the learning rate is 1.0 of the default of the optimization function. Further, the batch size was set to 50 and the number of iterations was 100. Experimentally, three BilSTA cells were selected, with hidden layers of sizes 100, 200, and 300, respectively.
Finally, three tasks were designed on the same data in order to demonstrate the effectiveness of the method of the invention. The first task is to use the multi-core CNN for relationship extraction, which can be regarded as a reference model; the second task is to use single-layer BilSTM to extract the relation, and experiments prove that the effect is superior to that of the simple multi-core CNN method through the combination of the BilSTM and the Attention; the third task is to use a Multi-BilSTMA model to carry out relationship extraction, prove that the model has the effect similar to the Multi-core CNN, can fully utilize the advantages of the BilSTM and the Attention, and obviously improve the experimental result compared with the former two.
After 5-fold cross-validation experiments, the properties are shown in table 1 (F values for the three models have been shown in bold).
TABLE 1 relationship extraction task Performance
The number distribution of each relationship type is not balanced, and the results are also directly shown in table 1. The overall appearance is that the result of the high number of types is also high, which also accords with the characteristics of the neural network. In general, the larger the data volume, the more sufficient the training, the less likely the overfitting, and the better the result, for the same data quality and the same model. It can also be seen from the results that the F values of the three classes "PART-white", "ORG-AFF" and "GEN-AFF" are significantly higher than those of the other three normal types, which is also determined by the large data volume of the three classes.
Meanwhile, as can be seen from table 1, the performance of single-layer BiLSTMA is superior to that of simple multi-core CNN, because BiLSTMA can capture dependency information and key features in sentences more effectively than CNN, thereby obtaining better extraction effect. And the Multi-BilSTMA has the characteristics of both, so the performance of the Multi-BilSTMA is obviously superior to that of the two. In conclusion, the Chinese relation extraction method based on the neural network provided by the invention has excellent performance.
The present invention is not described in detail, but is known to those skilled in the art. Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.
Claims (4)
1. The Chinese relation extraction method based on the neural network is characterized by comprising the following steps of:
step 1: constructing a BilSTMA unit, and extracting deep semantic information and global dependency information of a sentence;
step 2: constructing a Multi-BilSTMA model, and acquiring semantic information with dependency relationship of different granularities;
and step 3: the validity of the method is verified using the real data.
2. The method of claim 1, wherein step 1 constructs the BilSTMA unit to extract deep semantic information and dependency information of the sentence by fully utilizing the advantages of the two-way long-short term memory model in processing the long-term dependency problem and the feature that the attention mechanism can capture global dependency information.
3. The method of claim 1, wherein step 2, combining the BilSTA units with different sizes by setting hidden layers with different sizes in the BilSTA units, constructs a Multi-BilSTA model, and the model can obtain semantic information with dependency relationship with different granularities.
4. The method of claim 1, wherein step 3, the ACE RDC200 chinese dataset is used to verify the recognition and thus validity of the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669521.9A CN111008529B (en) | 2019-07-24 | 2019-07-24 | Chinese relation extraction method based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669521.9A CN111008529B (en) | 2019-07-24 | 2019-07-24 | Chinese relation extraction method based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111008529A true CN111008529A (en) | 2020-04-14 |
| CN111008529B CN111008529B (en) | 2023-07-21 |
Family
ID=70111470
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910669521.9A Active CN111008529B (en) | 2019-07-24 | 2019-07-24 | Chinese relation extraction method based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111008529B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111831783A (en) * | 2020-07-07 | 2020-10-27 | 北京北大软件工程股份有限公司 | Chapter-level relation extraction method |
| CN114647726A (en) * | 2022-03-04 | 2022-06-21 | 贵州大学 | News webpage information extraction method, system, equipment and medium based on multi-dimensional text features |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9799327B1 (en) * | 2016-02-26 | 2017-10-24 | Google Inc. | Speech recognition with attention-based recurrent neural networks |
| CN109299262A (en) * | 2018-10-09 | 2019-02-01 | 中山大学 | A kind of text implication relation recognition methods for merging more granular informations |
| US20190080225A1 (en) * | 2017-09-11 | 2019-03-14 | Tata Consultancy Services Limited | Bilstm-siamese network based classifier for identifying target class of queries and providing responses thereof |
| CN109710761A (en) * | 2018-12-21 | 2019-05-03 | 中国标准化研究院 | The sentiment analysis method of two-way LSTM model based on attention enhancing |
| CN109740148A (en) * | 2018-12-16 | 2019-05-10 | 北京工业大学 | A kind of text emotion analysis method of BiLSTM combination Attention mechanism |
| CN109858032A (en) * | 2019-02-14 | 2019-06-07 | 程淑玉 | Merge more granularity sentences interaction natural language inference model of Attention mechanism |
-
2019
- 2019-07-24 CN CN201910669521.9A patent/CN111008529B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9799327B1 (en) * | 2016-02-26 | 2017-10-24 | Google Inc. | Speech recognition with attention-based recurrent neural networks |
| US20190080225A1 (en) * | 2017-09-11 | 2019-03-14 | Tata Consultancy Services Limited | Bilstm-siamese network based classifier for identifying target class of queries and providing responses thereof |
| CN109299262A (en) * | 2018-10-09 | 2019-02-01 | 中山大学 | A kind of text implication relation recognition methods for merging more granular informations |
| CN109740148A (en) * | 2018-12-16 | 2019-05-10 | 北京工业大学 | A kind of text emotion analysis method of BiLSTM combination Attention mechanism |
| CN109710761A (en) * | 2018-12-21 | 2019-05-03 | 中国标准化研究院 | The sentiment analysis method of two-way LSTM model based on attention enhancing |
| CN109858032A (en) * | 2019-02-14 | 2019-06-07 | 程淑玉 | Merge more granularity sentences interaction natural language inference model of Attention mechanism |
Non-Patent Citations (1)
| Title |
|---|
| 程淑玉等: "融合Attention多粒度句子交互自然语言推理研究", 《小型微型计算机系统》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111831783A (en) * | 2020-07-07 | 2020-10-27 | 北京北大软件工程股份有限公司 | Chapter-level relation extraction method |
| CN111831783B (en) * | 2020-07-07 | 2023-12-08 | 北京北大软件工程股份有限公司 | Method for extracting chapter-level relation |
| CN114647726A (en) * | 2022-03-04 | 2022-06-21 | 贵州大学 | News webpage information extraction method, system, equipment and medium based on multi-dimensional text features |
| CN114647726B (en) * | 2022-03-04 | 2024-08-06 | 贵州大学 | News webpage information extraction method, system, equipment and medium based on multidimensional text features |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111008529B (en) | 2023-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109635109B (en) | Sentence classification method based on LSTM and combined with part-of-speech and multi-attention mechanism | |
| CN107133213B (en) | Method and system for automatically extracting text abstract based on algorithm | |
| WO2023065544A1 (en) | Intention classification method and apparatus, electronic device, and computer-readable storage medium | |
| CN112487203B (en) | Relation extraction system integrated with dynamic word vector | |
| CN107291693B (en) | Semantic calculation method for improved word vector model | |
| CN109783817B (en) | Text semantic similarity calculation model based on deep reinforcement learning | |
| US10394956B2 (en) | Methods, devices, and systems for constructing intelligent knowledge base | |
| CN106372061B (en) | Short text similarity calculation method based on semantics | |
| CN113255320A (en) | Entity relation extraction method and device based on syntax tree and graph attention machine mechanism | |
| CN106776562A (en) | A kind of keyword extracting method and extraction system | |
| CN110619051B (en) | Question sentence classification method, device, electronic equipment and storage medium | |
| CN111985228B (en) | Text keyword extraction method, text keyword extraction device, computer equipment and storage medium | |
| CN110347790B (en) | Text duplicate checking method, device and equipment based on attention mechanism and storage medium | |
| CN110765755A (en) | Semantic similarity feature extraction method based on double selection gates | |
| CN111680488A (en) | Cross-language entity alignment method based on knowledge graph multi-view information | |
| CN111475622A (en) | Text classification method, device, terminal and storage medium | |
| CN112232053A (en) | Text similarity calculation system, method and storage medium based on multi-keyword pair matching | |
| CN109918507B (en) | textCNN (text-based network communication network) improved text classification method | |
| CN114880461A (en) | Chinese news text summarization method combining contrast learning and pre-training technology | |
| CN111859979A (en) | Ironic text collaborative recognition method, ironic text collaborative recognition device, ironic text collaborative recognition equipment and computer readable medium | |
| CN110276396A (en) | Picture based on object conspicuousness and cross-module state fusion feature describes generation method | |
| CN113468854A (en) | Multi-document automatic abstract generation method | |
| CN111061873B (en) | Multi-channel text classification method based on Attention mechanism | |
| CN115422939A (en) | Fine-grained commodity named entity identification method based on big data | |
| CN111008529A (en) | Chinese relation extraction method based on neural network |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20241015 Address after: No. 001, Building 3, Computing Power Center, Guiyang Big Data Science and Technology Innovation City, Huchao Township, Gui'an New District, Guiyang City, Guizhou Province, 550000 Patentee after: Guizhou Luhao Technology Co.,Ltd. Country or region after: China Address before: Science and Technology Department of Huaxi north campus, Guizhou University, Huaxi District, Guiyang City, Guizhou Province Patentee before: Guizhou University Country or region before: China |