Ricoh Company, Ltd.

Boya Gongdao (Beijing) Robot Technology Co., Ltd.

Shanshan Jiang

Bin Dong

Yixuan Tong

Lei Ding

Yongwei Zhang

Jiashi Zhang

Yang Li

Rui Cheng

Liang Liang

Minglei Xiong

Yihan Li

Jian Wang

Jie Zhang

Jichuan Zheng

Lu Luo

Boyan Liu

Tianxiong Xiao

A method and an apparatus for named entity recognition, and a non-transitory computer-readable recording medium are provided. In the method, text elements are traversed according to a text span to obtain candidate entity words. Then, a class to which the candidate entity word belongs is recognized. The recognizing of the class includes generating a prompt template corresponding to the candidate entity word, and concatenating the text to be recognized and the prompt template to obtain a concatenated text; generating vector representations of the text elements in the concatenated text; generating the vector representation of the candidate entity word according to the vector representations of the text elements of each candidate entity word in the concatenated text, and the vector representation of the text element of the mask word; and classifying the vector representation of the candidate entity word to obtain the class of the candidate entity word.

Method and apparatus for named entity recognition, and non-transitory computer-readable recording medium

Disclosed are a table recognition method and apparatus. The table recognition method includes steps of obtaining an image vision feature and a character content feature of a table image; fusing the image vision feature and the character content feature of the table image to acquire a first fusion feature, and carrying out recognition based on the first fusion feature to acquire a table structure; and performing, based on the table structure, character recognition on the table image to acquire table character contents.

Table recognition method and apparatus and non-transitory computer-readable medium

Disclosed are a semantic matching and retrieval method and apparatus. The semantic matching and retrieval method includes steps of obtaining both the vector representation of a query text and the vector representation of a document text; obtaining the final vector representation of the query text; obtaining the final vector representation of the document text; calculating, based on the final vector representation of the query text and the final vector representation of the document text, the similarity score between the query text and the document text; and selecting, based on the similarity scores between the query text and a plurality of document texts, a document text matching the query text from the plurality of document texts.

Semantic matching and retrieval method and apparatus and non-transitory computer-readable medium

A method and an apparatus for machine reading comprehension, and a non-transitory computer-readable recording medium are provided. In the method, a paragraph-question pair is obtained, and subword vectors corresponding to subwords in the paragraph-question pair are generated. Then, for each subword, relative positions of the subword with respect to the other subwords are determined based on distances, and self-attention information of the subword in a first part and mutual attention information of the subword in a second part are calculated by using the relative positions and the subword vector. Then, a fusion vector of the subword is generated based on the self-attention information and the mutual attention information. Then, the fusion vectors of the subwords are input to a decoder of a machine reading comprehension model so as to obtain an answer predicted by the decoder.

Method and apparatus for machine reading comprehension, and non-transitory computer-readable recording medium

A method and an apparatus for sequence labeling on an entity text, and a non-transitory computer-readable recording medium are provided. In the method, a start position of an entity text within a target text is determined. Then, a first matrix is generated based on the start position of the entity text. Elements in the first matrix indicates focusable weights of each word with respect to other words in the target text. Then, a named entity recognition model is generated using the first matrix. The named entity recognition model is obtained by training using first training data, the first training data includes word embeddings corresponding to respective texts in a training text set, and the texts are texts whose entity label has been labeled. Then, the target text is input to the named entity recognition model, and probability distribution of the entity label is output.

Method and apparatus for sequence labeling on entity text, and non-transitory computer-readable recording medium

A method and an apparatus for fusing position information, and a non-transitory computer-readable recording medium are provided. In the method, words of an input sentence are segmented to obtain a first sequence of words in the input sentence, and absolute position information of the words in the first sequence is generated. Then, subwords of the words in the first sequence are segmented to obtain a second sequence including subwords, and position information of the subwords in the second sequence are generated, based on the absolute position information of the words in the first sequence, to which the respective subwords belong. Then, the position information of the subwords in the second sequence are fused into a self-attention model to perform model training or model prediction.

Method and apparatus for fusing position information, and non-transitory computer-readable recording medium

A method and an apparatus for recognizing an intention, and a non-transitory computer-readable recording medium are provided. The method includes learning vectors of knowledge base elements in corpus samples, and converting the corpus samples into row vectors composed of the vectors of the knowledge base elements in a knowledge base; extracting feature vectors from respective pooling windows in the corpus samples by hierarchical pooling, determining weights positively correlated with similarities between texts within the respective pooling windows and the respective corpus samples, performing weighting on the extracted feature vectors to obtain feature vectors of the respective pooling windows, and obtaining feature vectors of the respective corpus samples composed of the feature vectors of the pooling windows; training a vector-based intention recognition classifier, based on the feature vectors of the corpus samples; and recognizing an intention in querying a corpus, using the trained intention recognition classifier.

Method and apparatus for recognizing intention, and non-transitory computer-readable recording medium

An intention identification method includes generating a heterogeneous text network based on a language material sample; using a graph embedding algorithm to perform learning with respect to the heterogeneous text network and obtain a vector representation of the language material sample and a word, and determining keywords of the language material sample based on a similarity in terms of a vector between the language material sample and the word in the language material sample; training an intention identification model until a predetermined training termination condition is satisfied, by using the keywords of the language material samples, and obtaining the trained intention identification model; and receiving a language material query, and using the trained intention identification model to identify an intention of the language material query.

Intention identification method, intention identification apparatus, and computer-readable recording medium

Knowledge graph processing method and device are disclosed. The method includes steps of obtaining an entity set containing a first entity, a second entity, and relation information; acquiring text information and image information related to the first entity and the second entity; generating a first structural information vector of the first entity and a second structural information vector of the second entity, and creating a first text information vector of the first entity, a first image information vector of the first entity, a second text information vector of the second entity, and a second image information vector of the second entity; and building a joint loss function so as to attain a first target vector of the first entity, a second target vector of the second entity, and a target relation vector of the relation information between the first entity and the second entity.

Knowledge graph processing method and device

A method and an apparatus for processing word vectors of a neural machine translation model, and a non-transitory computer-readable recording medium are provided. In the method, word vectors that are input to an encoder and a decoder of a neural machine translation model are updated using semantic information among head representations at the same time and semantic information among head representations at different times, and the model is trained or translation is performed using the updated word vectors, thereby improving the model performance of the neural machine translation model.

Method and apparatus for processing word vector of neural machine translation model, and non-transitory computer-readable recording medium

A method and an apparatus for ranking responses of a dialog model, and a non-transitory computer-readable recording medium are provided. The dialog model is trained based on a sample data set. The method includes obtaining, from the sample data set, at least one similar dialog whose content is semantically similar to content of a target dialog; obtaining a probability of at least one target response generated by the dialog model when inputting the target dialog, and obtaining a probability of a target response generated by the dialog model when inputting the similar dialog; statistically analyzing, based on the probabilities of the respective generated target responses, scores of the target responses, the scores of the target responses being positively correlated with the probabilities of the target responses; and ranking the target responses in a descending order of the scores.

Method and apparatus for ranking responses of dialog model, and non-transitory computer-readable recording medium

A method, an apparatus and an electronic device for performing entity linking, and a non-transitory computer-readable recording medium are provided. The method includes constructing training data including a plurality of sets of labeled data using an existing unambiguous entity database where unambiguous entities corresponding to respective entity words are stored, each set of the labeled data including a text having an entity word and an unambiguous entity linked with the entity word; training an unambiguous entity recognition model whose output is a matching probability between an entity word in a text and an unambiguous entity using the training data; and inputting a text having an entity word to be recognized into the unambiguous entity recognition model, and determining an unambiguous entity linked with the entity word to be recognized based on an output result of the unambiguous entity recognition model.

Entity linking method, electronic device for performing entity linking, and non-transitory computer-readable recording medium

The single-jointed underwater robot fish includes a casing, a main board cabin, a motion control cabin and a battery cabin. The outer contour of the casing is a bionic fish shape, one end is a front casing, and the other end is a bionic fishtail structure. The front end of the main board cabin is affixed to the inner side of the front casing, a lens group, an eccentric anti-shake mechanism and a main PCB (printed circuit board) are set in the main board cabin. The motion control cabin is connected to the rear end of the main board cabin. A triaxial linkage device and a transmission device are set in the motion control cabin. The battery cabin is located below the motion control cabin.

Single-joint underwater robot fish

A method, an apparatus and a system for recognizing an evaluation element are provided. The method includes receiving an input text; performing, using a first conditional random field model, first recognition for the input text to obtain a first recognition result, the first recognition result including a pre-evaluation element that is recognized by using the first conditional random field model; performing, using a second conditional random field model, second recognition for the input text to obtain a second recognition result, the second recognition result including a false positive evaluation element that is recognized by using the second conditional random field model, the false positive evaluation element being an element erroneously detected as an evaluation element; and recognizing, based on the first recognition result and the second recognition result, an evaluation element in the input text.

Growing community of inventors

Beijing, China

Shanshan Jiang