Determining cross-document rhetorical connections based on parsing and identifying named entities

Abstract

An extended discourse tree and a method for navigating text using the extended discourse tree are provided. [Solution] A discourse navigation application creates a first discourse tree for a first paragraph of a first document and a second discourse tree for a second paragraph of a second document, determines entities and corresponding first basic discourse units from the first discourse tree, determines second basic discourse units in the second discourse tree that match the first basic discourse units, determines rhetorical connections between the two basic discourse units, and creates navigable links between the two discourse trees.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Application No. 62 / 564,961, filed September 28, 2017, and U.S. Provisional Application No. 62 / 729,335, filed September 10, 2018, which are incorporated herein by reference in their entireties.

[0002] Technical Field FIELD OF THE DISCLOSURE This disclosure relates generally to linguistics and, more particularly, to navigating one or more bodies of text using an extended discourse tree.

[0003] STATEMENT OF RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT Not applicable. [Background technology]

[0004] background Autonomous agents (chatbots) can be trained to answer user questions in a variety of applications, such as providing customer service. A variety of content can be used to train autonomous agents, such as customer service histories and other databases. However, such content may contain erroneous information, and an autonomous agent trained with that information may give incorrect answers to questions received from users.

[0005] Instead, formal documents can provide a higher level of accuracy. Examples of formal documents include banking process and procedure manuals. However, current analysis techniques, such as keyword-based search solutions, are unable to capture the relevance of various parts of these formal documents, leading to erroneous results. More specifically, keyword-based solutions that determine whether a keyword is present or absent cannot interpret dialogue (a series of related interactions) in the text. As a result, autonomous agents that rely on such solutions may not respond appropriately to user queries, reducing the effectiveness of such agents and causing user frustration.

[0006] Therefore, there is a need for improved solutions for searching and navigating within and between bodies of text. Summary of the Invention [Means for solving the problem]

[0007] Quick summary In general, the systems, devices, and methods of the present invention relate to extended discourse trees. In one example, a method accesses a first document and a second document. The method creates a first discourse tree for a first paragraph of the first document. The method creates a second discourse tree for a second paragraph of the second document. The method determines entities and corresponding first basic discourse units from the first discourse tree by extracting noun phrases from the discourse tree, classifying the noun phrases as entities or non-entities, and determining second basic discourse units in the second discourse tree that match the first basic discourse units. The method includes: In response to determining a rhetorical connection between the discourse unit and the second discourse unit, linking the first discourse tree and the second discourse tree via the rhetorical connection, thereby creating an extended discourse tree.

[0008] In one aspect, creating the first discourse tree and creating the second discourse tree further includes accessing sentences including fragments. At least one fragment includes a verb and words, each word including a word role within the fragment. Each fragment is a basic discourse unit. The creating further includes generating a discourse tree representing rhetorical connections between the multiple fragments. The discourse tree includes nodes, each non-terminal node representing a rhetorical connection between two fragments, and each terminal node of the discourse tree is associated with one of the multiple fragments.

[0009] In one aspect, the classifying includes one or more of using a trained machine learning model, a list of keywords, or searching internet resources.

[0010] In one aspect, an entity refers to either the name of a person, a company, a place, a document, or a date or time.

[0011] In one aspect, in response to not determining a rhetorical connection, the method creates a default rhetorical connection of type elaboration between the first basic discourse unit and the second basic discourse unit, and creates an extended discourse tree by linking the first discourse tree and the second discourse tree.

[0012] In one aspect, determining the rhetorical connection further includes combining the first basic discourse unit and the second basic discourse unit into a temporary paragraph and determining that the rhetorical connection is found within the temporary paragraph by applying discourse parsing to the temporary paragraph.

[0013] In one aspect, an entity is represented by either one or more phrases or one or more basic discourse units.

[0014] In one aspect, accessing the first document and the second document includes determining that a difference between a first content score of the first document and a second content score of the second document is within a threshold.

[0015] In one aspect, the first document and the second document are obtained by executing a user query of one or more documents.

[0016] In one aspect, the first document and the second document include text based on a particular topic.

[0017] In one aspect, accessing the first document and the second document includes determining that an existing link exists between the first document and the second document.

[0018] In a further aspect, a method for navigating a body of text using expanded discourse trees includes accessing an expanded discourse tree representing a document, the expanded discourse tree including a first discourse tree for a first document and a second discourse tree for a second document. The method further includes determining, from the expanded discourse tree, a first basic discourse unit responsive to a query from the user device and a first location corresponding to the first basic discourse unit. The method further includes determining, from the expanded discourse tree, a set of navigation options including a first rhetorical connection between the first base discourse unit and a second base discourse unit of the first discourse tree and a second rhetorical connection between the first base discourse unit and a third base discourse unit of the second discourse tree. The method further includes presenting the first rhetorical connection and the second rhetorical connection to the user device. The method further includes presenting the second base discourse unit to the user device in response to receiving a selection of the first rhetorical connection from the user device, or presenting the third base discourse unit to the user device in response to receiving a selection of the second rhetorical connection from the user device.

[0019] In one aspect, the method further includes, in response to receiving an additional query from the user device, determining additional basic discourse units responsive to the additional query and presenting the additional basic discourse units to the user device.

[0020] In one aspect, determining the first basic discourse unit further includes matching one or more keywords from the query in the first basic discourse unit.

[0021] In one aspect, determining the first basic discourse unit further includes generating a first parse tree for the query, generating additional parse trees for each of the one or more basic discourse units, and, in response to determining that one of the additional parse trees includes the first parse tree, selecting the basic discourse unit corresponding to the one additional parse tree as the first basic discourse unit.

[0022] In one aspect, the first rhetorical connection and the second rhetorical connection include one of an elaboration, an enablement, a condition, a contrast, or an attribution.

[0023] In one aspect, the above methods can be embodied on a tangible computer-readable medium and / or operate within a computer processor and associated memory.

[0024] In one aspect, a method determines rhetorical connections between one or more documents. The method accesses a first discourse tree representing a first document of a set of documents and a second discourse tree representing a second document from the set of documents. The method obtains a reference extended discourse tree from a set of extended discourse trees by applying the first discourse tree and the second discourse tree to a trained classification model. The trained classification model iterates through the set of extended discourse trees to identify a first candidate discourse tree and a second candidate discourse tree. The first candidate discourse tree and the second candidate discourse tree are best matches for the first discourse tree and the second discourse tree. The method determines one or more links between the first reference discourse tree and the second reference discourse tree from the reference extended discourse tree. The method propagates the one or more links to the first discourse tree and the second discourse tree, thereby creating an extended discourse tree.

[0025] In one aspect, the method further determines one or more rhetorical connections between the first discourse tree and the expanded tree based on the one or more links and presents the rhetorical connections on the user device. [Brief explanation of the drawings]

[0026] [Figure 1] 1 illustrates an exemplary rhetorical classification environment according to one aspect. [Figure 2] An example of a discourse tree with one aspect is shown below. [Figure 3] A further example of a discourse tree according to one aspect is shown below. [Figure 4] 1 illustrates an exemplary schema according to one aspect. [Figure 5] 1 illustrates a node-link representation of a hierarchical binary tree according to one aspect. [Figure 6] 6 illustrates an exemplary indented text encoding of the representation of FIG. 5, according to one aspect. [Figure 7] 1 illustrates an exemplary DT for an exemplary request regarding property taxes according to one aspect. [Figure 8] 8 shows exemplary responses to the questions presented in FIG. 7. [Figure 9] Shows the discourse tree of the official response according to one aspect. [Figure 10] 1 shows a discourse tree of raw answers according to one aspect. [Figure 11] 1 shows an example of an extended discourse tree according to one aspect. [Figure 12] 1 illustrates a flowchart of an example process for creating an extended discourse tree, according to one aspect. [Figure 13] This also shows, in one aspect, the relationships between text units of a document at different levels of granularity. [Figure 14] 1 illustrates a flowchart of an example process for navigating between documents using an extended discourse tree, according to one aspect. [Figure 15] 1 shows an example of an autonomous agent that answers a user's questions using an extended discourse tree, according to one aspect. [Figure 16] 1 shows an example of an extended discourse tree according to one aspect. [Figure 17] 1 illustrates a comparison between navigation using a search engine and navigation using an autonomous agent enabled with an extended discourse tree, according to one aspect. [Figure 18] 1 shows a simplified diagram of a distributed system for realizing one of the aspects. [Figure 19] FIG. 1 is a simplified block diagram of components of a system environment in which services provided by components of an aspect of the system may be offered as cloud services according to an aspect. [Figure 20] 1 illustrates an exemplary computer system in which various aspects of the present invention may be implemented. DETAILED DESCRIPTION OF THE INVENTION

[0027] Detailed Description As mentioned above, current keyword-based solutions are unable to capture the relevance of different pieces of text within a body of text, resulting in autonomous agents that attempt to mimic human dialogue without a full understanding of the communicative discourse. Such solutions result in the autonomous agents issuing a random series of utterances, making it difficult to accomplish tasks or provide recommendations.

[0028] In contrast, aspects disclosed herein provide technical improvements in the areas of computer-implemented linguistics and text navigation. More specifically, certain aspects create an extended discourse tree that represents not only the rhetorical connections between discourse units in a particular document, but also the rhetorical connections between identified entities across multiple documents. In doing so, certain aspects provide a cohesive discourse flow for interacting with an autonomous agent or for searching and navigating a body of text organized or divided into different documents.

[0029] For example, a discourse navigation application might build individual discourse trees for various text units (such as paragraphs), perform discourse analysis to determine rhetorical relationships between the discourse trees, and then create a single extended discourse tree from them. The extended discourse tree includes inter-document rhetorical relationships in addition to the intra-document rhetorical relationships within the individual discourse trees. The discourse navigation application can then use the extended discourse tree to facilitate autonomous agents or search.

[0030] Discourse trees originate from the Theory of Rhetorical Structure (RST). RST models the logical organization of text used by writers, relying on the relationships between parts of the text. RST simulates text coherence by forming a hierarchical, connected structure of the text through a discourse tree. Rhetorical relationships are divided into equivalence classes and subclasses; these relationships hold across two or more text spans, thus achieving coherence. These text spans are called elementary discourse units (EDUs). Clauses within a sentence and sentences within a text are logically connected by the author. The meaning of a given sentence is related to the meaning of the previous and following sentences.

[0031] The leaves of a discourse tree correspond to EDUs, consecutive atomic text spans. Adjacent EDUs are connected by coherence relations (such as attribution, sequence, etc.) to form higher-level discourse units. The leaves of a particular EDU are logically related. This relationship is called the coherence structure of the text. Examples of relationships include elaboration and enablement. As used herein, "coreness" refers to which text segment, fragment, or span is more central to the writer's purpose. "Core" refers to a span of text that is more central to the writer's purpose than "satellites" that are less central to the writer's purpose.

[0032] The following non-limiting example is provided to introduce particular aspects: A discourse navigation application running on a computing device accesses a set of input documents. The discourse navigation application creates an individual discourse tree for each paragraph in each document. Each discourse tree identifies rhetorical relationships between entities, thereby facilitating navigation between entities (e.g., topics such as places, things, and people). For example, using discourse trees, a user can navigate from a text describing California to a text containing information about California cities such as San Francisco and a text that further describes that entity (San Francisco) based on the identification of the common entity (California).

[0033] In addition to determining rhetorical relationships within a particular discourse tree, the discourse navigation application performs additional discourse analysis between discourse trees and creates links between documents based on the analysis. More specifically, the discourse navigation application identifies entities in a first discourse tree (e.g., representing a first document) and identifies any such entities in a second discourse tree (e.g., representing a second document), and then determines rhetorical connections between the corresponding entities. In this manner, the discourse navigation application enables navigation between entities represented in multiple documents.

[0034] For example, a discourse navigation application might relate a paragraph in a first document that discusses baseball to a second paragraph in a second document that provides additional information and elaborates on baseball, a third paragraph in a third document that gives examples of baseball teams such as the Atlanta Braves, and so on.

[0035] In another example, a discourse navigation application can identify rhetorical relations of contrast and provide contrasts to the user. For example, an entity in a first document takes a position on a topic, e.g., "Some experts believe that climate change is caused by human activity," contrasted with text related to the entity in a second document, "However, a minority of experts believe that climate change is cyclical." The extended discourse tree can represent multiple rhetorical relations between documents, including context, justification, motivation, etc.

[0036] In certain aspects, the trained machine learning model is used to construct additional augmented discourse trees. The trained classification model can create an augmented discourse tree from a discourse tree for text in a first domain using a set of augmented discourse trees for text in a second domain. For example, an augmented discourse tree can be created from multiple discourse trees from legal documents using a classification model trained on augmented discourse trees from current events.

[0037] Specific Definitions As used herein, a "text unit" refers to a unit of text. Examples include a basic discourse unit, a phrase, a fragment, a sentence, a paragraph, a page, and a document.

[0038] As used herein, an "entity" refers to something with a distinct and independent existence. Entities can be used in text units. Examples of entities include the names of people, companies, places, documents, or dates or times.

[0039] As used herein, "rhetorical structures theory" is a field of research and study that has provided a theoretical foundation upon which the coherence of discourse can be analyzed.

[0040] As used herein, "discourse tree" or "DT" refers to a structure that expresses rhetorical relationships about sentences or parts of sentences.

[0041] As used herein, "rhetorical relation" and "rhetorical connection" are A rhetorical relationship, coherence relationship, or discourse relationship describes how two segments of discourse are logically connected to each other. Examples of rhetorical relationships include elaboration, contrast, and attribution.

[0042] As used herein, a "sentence fragment" or "fragment" is a part of a sentence that can be separated from the rest of the sentence. A fragment is a basic discourse unit. For example, in the sentence "Dutch accident investigators say that evidence points to pro-Russian rebels as being responsible for shooting down the plane," the two fragments are "Dutch accident investigators say that evidence points to pro-Russian rebels" and and "as being responsible for shooting down the plane." Fragments may contain verbs, but do not necessarily have to.

[0043] As used herein, an "index" is a table, data structure, pointer, or other mechanism that links two keywords, pieces of data, or pieces of text. An index can contain searchable content. Examples of indexes include inverted indexes, searchable indexes, and string matches. Inverted indexes are also searchable.

[0044] Referring now to the figures, Figure 1 illustrates an exemplary document navigation environment according to one aspect. Figure 1 illustrates input documents 110a-n, a computing device 101, a user device 170, and a data network 104.

[0045] The computing device 101 includes a discourse navigation application 102, The computing device includes an expanded discourse tree 130, a machine learning model 120, and training data 125. In one example, the discourse navigation application 102 receives input documents 110a-n, creates a discourse tree for each input document 110a-n, determines entities within the generated discourse tree, determines rhetorical relationships between the entities, and creates the expanded discourse tree 130. The computing device can also receive queries from a user device 170 and process those queries by navigating the expanded discourse tree 130.

[0046] User device 170 may be any mobile device, such as a mobile phone, smartphone, tablet, laptop, or smartwatch. User device 170 communicates with computing device 101 or a remote server via data network 104. Data network 104 may be any public or private network, wired or wireless network, wide area network, local area network, or the Internet. The functionality of user device 170 may be implemented in software, for example, via an application or web application. User device 170 includes display 171. Examples of display 171 include a computer screen, a mobile device screen, an LCD, or an LED-backlit display, etc.

[0047] Display 171 shows three messages 181-183. Continuing with the example, discourse navigation application 102 receives message 181 from user device 170. Message 181 is a user query: "Can I use one credit card to pay for another?" Discourse navigation application 102 accesses extended discourse tree 130 and determines that entity "credit card" exists in the discourse tree of the first document. Discourse navigation application 102 then determines that several different possibilities exist for further elaborating on entity "credit card." Specifically, discourse navigation application 102 determines that the noun phrases "balance transfer," "draw funds from checking account," and "cancel your credit card" are each linked from entity "credit card" and are found in one or more other discourse trees within extended discourse tree 130. Accordingly, discourse navigation application 102 presents message 182 on the user device. In response, if the user selects "cancel," as indicated by message 183, then the discourse navigation application follows the rhetorical relationship between the entities "credit card" and "cancel your credit card."

[0048] The discourse navigation application 102 can continue this process, facilitating convergence in this answer navigation session because it can suggest additional answers based on additional explanations provided after the message 182 is read by the user. The additional answers or text can be suggested based on specific rhetorical relationships. For example, presenting text related by an elaboration relationship helps the user elaborate on a topic, while presenting text related by a contrast relationship provides contrast for the user. In this way, the suggested topics provide the user with an opportunity to evaluate how the request was understood and learn some basic knowledge associated with the question, resulting in improved convergence over traditional search engine-based solutions. For example, the document navigation options 102 can suggest options such as "bad judgment," "isolation," or "out of economic range."

[0049] Rhetorical Structure Theory and Discourse Trees Linguistics is the scientific study of language. For example, linguistics can include the structure of a sentence (syntax), e.g., subject-verb-object, the meaning of a sentence (semantics), e.g., "dog bites man" vs. "man bites dog", and also what speakers do in conversation, i.e., discourse analysis or the analysis of language beyond the sentence.

[0050] The theoretical foundations of discourse (Rhetoric Structure Theory (RST)) are found in Mann, William and Thompson, Sandra, "Rhetorical structure theory: A Theory of Text-Interdisciplinary Journal for the Study of Discourse )8(3):243-281: 1988). RST enabled the analysis of discourse in a similar way to how the syntax and semantics of programming language theory enabled modern software compilers. More specifically, RST envisions building blocks at at least two levels: a first level, such as kernelity and rhetorical relations, and a second level, such as structure or schema. A discourse parser or other computer software can parse text into a discourse tree.

[0051] Rhetorical relations As mentioned above, several aspects described in this specification use rhetorical relations and discourse trees. Rhetorical relations can be described in various ways. For example, Mann and Thompson describe 23 possible relations. There is "Rhetorical Structure Theory: A Theory of Text Organization" by C. Mann, William & Thompson, Sandra (1987) ("Mann and Thompson"). Several other relations are also possible. Table 2 below lists the different rhetorical relations.

[0052] [Table 1]

[0053] Some empirical studies have assumed that the majority of texts are constructed using core-satellite relationships (see Mann and Thompson). However, other relationships do not involve a finite selection of cores. Examples of such relationships are shown in Table 3 below.

[0054] [Table 2]

[0055] Figure 2 shows an example of a discourse tree according to one aspect. Figure 2 includes discourse tree 200. The discourse tree includes text span 201, text span 202, text span 203, relation 210, and relation 211. The numbers in Figure 2 correspond to the three text spans. Figure 2 corresponds to the following text example with three text spans numbered 1, 2, and 3:

[0056] 1. Honolulu, Hawaii will be the site of the 2017 Conference on Hawaiian History.

[0057] 2. It is expected that 200 historians from the US and Asia will attend.

[0058] 3. The conference was about how Polynesians sailed to Hawaii. will be concerned with how the Polynesians sailed to Hawaii).

[0059] For example, relationship 210, or elaboration, describes the relationship between text span 201 and text span 202. Relationship 210 indicates the relationship (elaboration) between text span 203 and text span 204. As shown, text spans 202 and 203 further elaborate on text span 201. In the example above, assuming the purpose is to inform the reader of a conference, text span 1 is the core. Text spans 2 and 3 provide more details about the conference. In Figure 2, horizontal numbers (e.g., 1-3, 1, 2, 3) cover spans of text (possibly composed of further spans), and vertical lines indicate a core or multiple cores. The curved lines represent rhetorical relationships (elaborations), with the arrows pointing from the satellites to the core. If only text spans functioned as satellites, rather than as cores, the text would remain coherent even if the satellites were removed. Removing the core from Figure 2 would make text spans 2 and 3 difficult to understand.

[0060] Figure 3 shows a further example of a discourse tree according to one aspect. Figure 3 includes constituents 301 and 302, text spans 305-307, relation 310, and relation 311. Relation 310 enables describing the relationship between constituent 306 and constituent 305, and between constituent 307 and constituent 305. Figure 3 points to the following text spans: 1. The new Tech Report abstracts are now in the journal area of ​​the library near the abridged dictionary.

[0061] 2. Please sign your name by any means that you would be interested in seeing.

[0062] 3. The last day for sign-ups is May 31st. As can be seen, relationship 310 shows the relationship, or enabling, between entity 307 and entity 306. Figure 3 illustrates that although multiple kernels can be nested, there is only one most kernel text span.

[0063] Discourse tree construction Discourse trees can be generated using a variety of methods. A simple example of a method for building a DT bottom up is as follows: (1) Divide the discourse text into multiple units using (a) and (b) below.

[0064] (a) The unit size may vary depending on the purpose of the analysis. (b) Typically the unit is a clause.

[0065] (2) Examine each unit and each of its neighboring units. Is there a relationship between them? (3) If the relationship is maintained, mark the relationship.

[0066] (4) If a relationship does not hold, the unit may be at the boundary of a higher-level relationship. Focus on the relationships that hold between larger units (spans).

[0067] (5) Continue until all units in the text are understood. Mann and Thompson also describe a second level of building block structures called schema applications. In RST, rhetorical relations are not mapped directly onto the text; rather, they are fitted onto structures called schema applications, which are then further fitted to the text. Schema applications are derived from simpler structures called schemas (as shown in Figure 4). Each schema shows how a particular unit of text can be decomposed into smaller textual units. A rhetorical structure tree, or DT, is a hierarchical system of schema applications. Schema applications link several consecutive text spans to create complex text spans. Complex text spans can then be linked by higher-level schema applications. RST asserts that the structure of any coherent discourse can be described by a single rhetorical structure tree, the top-level schema of which creates a span that encompasses the entire discourse.

[0068] FIG. 4 illustrates an exemplary schema according to one aspect. FIG. 4 illustrates that a joint schema is a list of items consisting of a core but no satellites. FIG. 4 illustrates schemas 401-406. Schema 401 illustrates a situational relationship between text span 410 and text span 411. Schema 402 illustrates a sequence relationship between text span 420 and text span 421, and a sequence relationship between text span 421 and text span 422. Schema 403 illustrates a contrastive relationship between text span 430 and text span 431. Schema 404 illustrates a joint relationship between text span 440 and text span 441. Schema 405 illustrates a motivational relationship between 450 and 451, and an enabling relationship between 452 and 451. Schema 406 illustrates a joint relationship between text span 460 and text span 462. An example of a joint scheme is illustrated in FIG. 4 for the following three text spans:

[0069] 1. Skies will be partly sunny in the New York metropolitan area today.

[0070] 2. It will be more humid, with temperatures in the middle 80's.

[0071] 3. Tonight will be mostly cloudy, with the low temperature between 65 and 70 degrees Fahrenheit.

[0072] Although Figures 2-4 show discourse trees in several graphs, other representations are possible.

[0073] Figure 5 shows a node-link representation of a hierarchical binary tree according to one aspect. As can be seen from Figure 5, the leaves of the DT correspond to consecutive but non-overlapping text spans called elementary discourse units (EDUs). Adjacent EDUs are connected by relations (e.g., elaboration, attribution, etc.) and form larger discourse units connected by relations. "Discourse analysis in RST involves two subtasks: discourse segmentation is the task of identifying EDUs, and discourse parsing is the task of linking discourse units into a labeled tree." See Joty, Shafiq R, Giuseppe Carenini, Raymond T Ng, and Yashar Mehdad (2013), "Combining intra- and multi-sentential rhetorical parsing for document-level discourse analysis," ACL (1), pages 486-496. I want to be.

[0074] Figure 5 shows text spans that are leaves or terminal nodes on a tree, numbered in the order in which they appear throughout the text shown in Figure 6. Figure 5 includes a tree 500. Tree 500 includes, for example, nodes 501-507. The nodes represent relationships. The nodes are either non-terminal nodes, such as node 501, or terminal nodes, such as nodes 502-507. As can be seen, nodes 503 and 504 are related by a joint relationship. Nodes 502, 505, 506, and 508 are nuclei. The dotted lines indicate that the branches or text spans are satellites. These relationships are the nodes in the gray boxes.

[0075] FIG. 6 shows an exemplary indented text encoding for the representation in FIG. 5, according to one aspect. FIG. 6 includes text 600 and text sequences 602-604. Text 600 is represented in a manner that is more amenable to computer programming. Text sequence 602 corresponds to node 502. Sequence 603 corresponds to node 503. Sequence 604 corresponds to node 504. In FIG. 6, "N" indicates a nucleus and "S" indicates a satellite.

[0076] Discourse Parser Example Automatic discourse segmentation can be performed in various ways. For example, given a sentence, a segmentation model identifies boundaries of composite basic discourse units by predicting whether a boundary should be inserted before each particular token in the sentence. For example, one framework considers each token in the sentence sequentially and independently. In this framework, the segmentation model scans the sentence token by token and uses binary classification, such as a support vector machine or logistic regression, to predict whether it is appropriate to insert a boundary before the token being examined. In another example, the task is a sequential labeling problem. Once the text is segmented into basic discourse units, a sentence-level discourse analysis can be performed to construct a discourse tree. Machine learning techniques can be used.

[0077] In one aspect of the present invention, two Rhetorical Structure Theory (RST) discourse parsers are used: CoreNLPProcessor, which relies on constituent syntax, and FastNLPProcessor, which uses dependency syntax. Refer to "Two Practical Rhetorical Structure Theory Parsers" (2015) stomach.

[0078] In addition, the two discourse parsers mentioned above, i.e., CoreNLPProcessor and FastNLPProcessor, use Natural Language Processing (NLP) for parsing. For example, Stanford CoreNLP presents basic shapes for multi-word parts of speech, whether they are names of companies, people, etc.; standardizes dates, times, and numerical quantities; marks sentence structure in terms of phrases and syntactic dependencies; and indicates which noun phrases refer to the same entity. In practice, RST remains a theory of discourse that may work in many cases, but may not work in other cases. There are many variables, including but not limited to what EDUs are in a coherent text, i.e., what discourse segmenter is used, what relation inventory is used, what relations are selected for the EDUs, the corpus of documents used for training and testing, and even what parser is used. Thus, for example, in the aforementioned paper "Two Practical Rhetorical Structure Theory Parsers" by Surdeanu et al., specialized parsers are used to determine which parser will give better performance. Testing must be performed on a specific corpus using the metrics determined. Thus, unlike computer language parsers, which produce predictable results, discourse parsers (and segmenters) can produce unpredictable results depending on the training and / or testing text corpus. Discourse trees are thus a mixture of predictable techniques (e.g., compilers) and unpredictable techniques (e.g., chemistry, where experimentation is required to determine which combinations will produce the desired results).

[0079] To objectively judge how good a discourse analysis is, for example, read Daniel Marcu's "The Theory and Practice of Discourse Parsing and Summarization" (MIT Press) (2000) and a series of other metrics such as Precision / Recall / F1. Metrics are used. Precision, or positive predictive value, is the fraction of informative instances among retrieved instances, while recall (also known as sensitivity) is the fraction of informative instances retrieved over the total amount of informative instances. Thus, both precision and recall are based on understandings and criteria of relevance. Suppose a computer program for recognizing dogs in photos identifies eight dogs in a photo containing twelve dogs and some cats. Of the eight dogs identified, five are actually dogs (true positives) and the rest are cats (false positives). The program's precision is 5 / 8 and its recall is 5 / 12. If a search engine returns 30 pages, but only 20 of them are informative and does not return 40 additional informative pages, its precision is 20 / 30 = 2 / 3 and its recall is 20 / 60 = 1 / 3. Thus, in this case, precision is "how useful the search results are" and recall is "how complete the results are." The F1 score (F-score or F-criterion) is a measure of the accuracy of a test. It takes into account both the precision and recall of a test to calculate the score. F1=2x(precision x recall) / (precision + recall)), which is the harmonic mean of precision and recall. The F1 score reaches its optimum at 1 (perfect precision and recall) and its worst at 0.

[0080] Analyzing request and response pairs 7 shows an exemplary discourse tree for the property tax request example, according to one aspect. Node labels are relationships, and arrowhead lines point to satellites. Cores are solid lines. FIG. 7 shows a discourse tree 700 representing the following text:

[0081] Request: "My husbands' grandmother gave him his grandfather's truck. She signed the title over but due to my husband having unpaid fines on his license, he was not able to get the truck put in his name. I wanted to put in my name and paid the property tax and got insurance for the truck. By the time it came to sending off the title and getting the tag, I didn't have the money to do so. Now, due to Due to certain circumstances, I am not going to be able to afford the truck. I went to the insurance company and was refused a refund. I am just wondering that since I am not going to have a tag on this truck, is it possible to get the property tax refunded? (My husband's grandmother gave him his grandfather's truck. She transferred the ownership.) However, my husband had outstanding fines on his license so he couldn't put the truck in his name. I wanted it in my name, so I paid the property taxes and got insurance on the truck. When it came time to get the title and get the identification tag, I didn't have the money to do so. Now, depending on the situation, I won't be able to afford to keep the truck. I went to my insurance agent and they refused to refund me. I'm wondering if I can get the property taxes refunded if I can't get the identification tag on this truck."

[0082] Response: "The property tax is assessed on property that you own. Just because you chose to not register it does not mean that you don't own it, so the tax is not refundable. Even if you have not titled the vehicle yet, you still own it within the boundaries of the tax district, so the tax is payable. Note that all states give you a limited amount of time to transfer title and pay the use tax. If you apply late, there will be penalties on top of the normal taxes and fees. You don't need to register it at the same time, but you absolutely need to title it within the period of time stipulated in state law. Tax is levied on the property of the vehicle. Just because you choose not to register it does not mean you do not own it. Therefore, the tax cannot be refunded. Even if you have not yet acquired title to the vehicle, you still own the vehicle within the tax district boundaries and must pay the tax. Keep in mind that every state has a limited time to transfer title and pay the use tax. If you file late, you may be assessed penalties in addition to the regular taxes and fees. You do not have to register the vehicle at the same time, but you must ensure you acquire title within the time frame required by state law. As can be seen from Figure 7, by analyzing the above text, we get the following results: "My husbands' grandmother gave him his grandfather's truck" is a translation of "I wanted to This is detailed by the phrase "She signed the title over but due to my husband", which is detailed by the phrases "having unpaid fines on his license, he was not able to get the truck put in his name", "and paid the property tax", and "and got insurance for the truck".

[0083] "My husband's grandmother gave him his grandfather's truck. She signed the title over, but due to my husband having unpaid fines on his license, he was unable to get the truck in his name. I wanted to put it in my name, paid the property tax, and got insurance for the truck." is detailed by "to do so," which contrasts with "By the time," which is detailed by "it came to sending off the title." Elaborated by "I didn't have the money."

[0084] "My husbands' grandmother gave him his grandfather's truck. She signed the title over but due to my husband having unpaid fines on his license, he was not able to get the truck put in his name. I wanted to put in my name and paid the property tax and got insurance for the truck. By the time it came to sending off th e title and getting the tag, I didn't have the money to do so" is contrasted with "Now, due to circumstances," which is detailed by "I am not going to be able to afford the truck," which is detailed by "I went to the insurance place" and "and was refused a refund."

[0085] "My husbands' grandmother gave him his grandfather's truck. She signed the title over but due to my husband having unpaid fines on his license, he was not able to get the truck put in his name. I wanted to put in my name and paid the property tax and got insurance for the truck. By the time it came to sending off the title and getting the tag, I didn't have the money to do so. Now, due to circumstances, I am not going to be able to afford the truck. I went to the insurance place and was refused a refund." is detailed in "I am just wondering that since I am not going to have a tag on this truck, is it possible to get the property tax refunded?"

[0086] "I am just wondering" is the same as "Is it possible to get the property tax refunded?" with the condition "since I am not going to have a tag on this truck." It belongs to the unit "that".

[0087] As can be seen, the main topic theme is "Property Tax on Cars." The question contains a contradiction: on the one hand, all property is taxable; on the other hand, ownership is somewhat incomplete. A good response must address the topic of the question and clarify the contradiction. To do this, the respondent makes a stronger claim about the need to pay tax on everything owned, regardless of registration status. This example is an element of a positive training set from the Yahoo! Answers reputation domain. The main topic theme is "Property Tax on Cars." The question contains a contradiction: on the one hand, all property is taxable; on the other hand, ownership is somewhat incomplete. A good answer / response must address the topic of the question and clarify the contradiction. The reader may realize that the question contains a rhetorical relation of contrast, and therefore the answer must match the question in a similar relation to be convincing. In other cases, the answer may appear incomplete even to people who are not experts in the field.

[0088] 8 illustrates an exemplary response to the question presented in FIG. 7, in accordance with certain aspects of the present invention. FIG. 8 illustrates a discourse tree 800. The central kernel is "The property tax is assessed on property," which is elaborated by "that you own." "The property tax is assessed on property that you own" also translates to "Just because you chose not to register." it does not mean that you don't own it, so the tax is not refundable. Even if you have not titled the vehicle yet, you still own it within the boundaries of the tax district, so the tax is payable. Note that all states give you a limited amount of time to transfer title and pay the use tax. .

[0089] At its core: "The property tax is assessed on property that you own. Just because you chose to not register it does not mean that you don't own it, so the tax is not refundable. Even if you have not titled the vehicle yet, you still own it within the boundaries of the tax district, so the tax is payable. Note that all states give you a limited amount of time to transfer title and pay the use tax." This is detailed by the condition, "If you apply late," that "there will be penalties on top of the normal taxes and fees." This is further explained by the condition, "but you absolutely This is elaborated by the contrast between "You need to title it within the period of time stipulated in state law" and "You don't need to register it at the same time."

[0090] By comparing the DT of Figure 7 with the DT of Figure 8, it is possible to determine how well the response (Figure 8) matches the request (Figure 7). In some aspects of the invention, the above framework is used, at least in part, to determine the DT for request / response and rhetorical fit between the DTs.

[0091] In another example, the question "What does The Investigative Committee of the Russian Federation do?" has at least two answers, for example, the official answer or the actual answer.

[0092] Figure 9 shows a discourse tree for an official response according to one aspect. Figure 9 is a discourse tree 900 for an official response or statement that states, "The Investigative Committee of the Russian Federation is the main federal investigating authority which operates as Russia's Anti-corruption agency and has statutory responsibility for inspecting the police forces, combating police corruption and police misconduct, is responsible for conducting investigations into local authorities and federal governmental bodies."

[0093] 10 shows a discourse tree 1000 for raw answers according to one aspect. As shown in FIG. 10, an alternative, possibly more honest answer is as follows: "The Investigative Committee of the Russian Federation is supposed to fight corruption. However, top-ranking officers of the Investigative Committee of the Russian Federation are charged with the creation of a criminal community. Not only that, but their involvement in large bribes, money laundering, obstruction of justice, abuse of power, extortion, and racketeering has been reported. Due to the activities of these officers, dozens of high-profile cases, including those against criminal lords, had been ultimately ruined." The choice of answer depends on the context. Rhetorical structures include "official", "politically correct" template-based answers and "actual", " "raw," "reports from the field," or "controversial" This allows us to distinguish between "controversial" and "controversial" responses (see Figures 9 and 10). (See references below.) Sometimes the question itself can give a hint as to which category of answer is expected. If the question is formulated as a factual or definitional question without a second meaning, an answer in the first category is appropriate. In other cases, if the question has the meaning "tell me what it actually is," the second category is appropriate. In general, after extracting the rhetorical structure from the question, it is easier to select an appropriate answer that will have a similar, matching, or complementary rhetorical structure.

[0094] The official response is a clarification and comment that is neutral in terms of the debate the text may contain. It is based on joints (see Figure 9). At the same time, the raw answers contain contrastive relations: this relation between phrases about what the agent is expected to do and phrases about what this agent was found to have done is extracted.

[0095] Extended Discourse Tree Aspects of the present disclosure facilitate navigating an extended discourse tree constructed from a corpus of related content, such as multiple documents. An extended discourse tree is a combination of discourse trees consisting of individual text units (e.g., paragraphs) from multiple documents. In various aspects, the extended discourse tree is used to enable not only zooming in based on keywords, but also navigation in, out, or back based on how documents are interconnected, thereby enabling an autonomous agent to provide content navigation, such as guided search.

[0096] FIG. 11 shows an example of an extended discourse tree according to one aspect. 100. Expanded discourse tree 1100 includes groups 1100, 1120, 1130, 1140, and 1150. Each group includes a document and a discourse tree generated from the document. For example, group 1110 includes discourse tree 1111 and document 1112, group 1120 includes discourse tree 1121 and document 1122, and so on.

[0097] In addition to links within a particular discourse tree, e.g., between discourse trees 1111, 1121, 1131, 1141, and 1151, the extended discourse tree 1100 includes inter-discourse tree links 1161-1164 and associated inter-document links 1171-1174. As will be further described with respect to Figure 12, the discourse navigation application 102 constructs discourse trees 1111-1115. Discourse tree 1111 represents document 1112, discourse tree 1121 represents document 1122, and so on. The extended discourse tree 1100 is constructed by building a discourse tree for each paragraph or document.

[0098] Inter-discourse tree link 1161 connects discourse trees 1111 and 1121, inter-discourse tree link 1162 connects discourse trees 1121 and 1131, inter-discourse tree link 1163 connects discourse trees 1111 and 1141, and inter-discourse tree link 1164 connects discourse trees 1121 and 1151. Based on the inter-discourse tree links 1161-1164, the discourse navigation application 102 creates inter-document links 1171, 1172, 1173, and 1174 corresponding to inter-discourse tree links 1161, 1162, 1163, and 1164, respectively. Inter-document links 1171-1174 can be used to navigate documents 1112, 1122, 1132, 1142, and 1152.

[0099] The discourse navigation application 102 determines one or more entities in a first one of the discourse trees 1111-1115. Examples of entities include places, things, people, or businesses. The discourse navigation application 102 then identifies the same entities that exist in other discourse trees. Based on the determined entities, the discourse navigation application 102 determines rhetorical connections between each matching entity.

[0100] For example, "San Francisco is in California" and the entity "San Francisco" appears in document 1112 and Statement 1122 further explains that "San Francisco has a moderate climate but can be quite windy." In this case, the discourse navigation application 102 may is one of "elaboration," and would mark links 1161 and 1171 as "elaboration." Continuing the example, the discourse navigation application 102 determines links 1162-1164 and corresponding links 1172-1174 based on the determined rhetorical relationship. The discourse navigation application 102 combines the discourse trees of the paragraphs of the document to form an extended discourse tree 1100.

[0101] Using the links in the expanded discourse tree 1100, a discourse navigation application can navigate between paragraphs of the same document or between documents, such as documents 1112 and 1122. For example, if a user is interested in more information about a particular topic, the discourse navigation application 102 navigates through nucleus-to-satellite elaboration rhetorical relations within a paragraph or through elaboration rhetorical relation hyperlinks to documents that provide more specific information about the topic.

[0102] Conversely, if the user determines that the suggested topic is not exactly what is needed, the user can return to a higher-level view of the document (e.g., from satellite to core, or from narrow document to broader document). The discourse navigation application 102 then navigates the elaboration relationships in reverse order, from satellite to core, either paragraph-by-paragraph or between documents. Similarly, the discourse navigation application 102 facilitates other navigation options, such as relying on contrast or conditional rhetorical connections to explore contentious topics.

[0103] The discourse navigation application 102 identifies inter-entity relationships using fictitious text fragments or temporary paragraphs from each text fragment of the original paragraph to build rhetorical links between text fragments of different paragraphs or documents, and performs coreference analysis and discourse syntactic analysis on the paragraphs.

[0104] 12 shows a flowchart of an example process 1200 for creating an expanded discourse tree, according to one aspect. The input of process 1200 is a set of documents, and the output is an expanded discourse tree, which is encoded as a regular discourse tree with document-identifying labels for each node. For illustrative purposes, process 1200 is described with respect to two documents, e.g., documents 110a-b, although process 1200 can be used with any number of documents.

[0105] At block 1201, process 1200 includes accessing a first document and a second document. Examples of documents include texts, books, news articles, and other electronic documents.

[0106] In one aspect, the discourse navigation application 102 selects documents that are topically similar or identical. For example, the discourse navigation application 102 can determine a content score for each document, e.g., by determining the similarity of keywords between the documents. For example, the discourse navigation application 102 determines that a first content score of a first document and a second content score of a second document are within a threshold, and creates an expanded discourse tree using the first and second documents based on the similarity.

[0107] In one aspect, the discourse navigation application 102 performs document analysis, including generating a document tree that represents the sentence and phrase structure of the document. The rhetorical relations associated with inter-document links can determine various navigation scenarios. By default, elaboration can be used. The discourse navigation application 102 provides links to other documents related by attribution when the user is interested in questions such as "why" or "how." The discourse navigation application 102 can provide links to documents related by contrast when the user expresses disagreement with the initially presented document or requests a document that provides a contrast to the current document.

[0108] In a further aspect, the discourse navigation application 102 retrieves the first and second documents by executing a user query, examples of which include "climate change" and "linguistics documents."

[0109] At block 1202, the process 1200 includes creating a first discourse tree for a first paragraph of a first document. The discourse navigation application 102 accesses a paragraph from the first document. Each sentence of the paragraph includes a fragment, or basic discourse unit. At least one fragment includes a verb. Each word in the fragment includes its role (e.g., function, etc.) in the fragment. The discourse navigation application 102 generates a discourse tree that represents the rhetorical connections between the fragments. The discourse tree includes multiple nodes, each non-terminal node representing a rhetorical connection between two fragments, and each terminal node is associated with one of the multiple fragments. The discourse navigation application 102 continues in this manner, building a set of discourse trees for each paragraph of the first document. Although the process 1200 is described with respect to a paragraph as a unit of text, other sizes of text can also be used.

[0110] At block 1203, process 1200 includes creating a second discourse tree for a second paragraph of the second document. At block 1203, process 1200 performs substantially similar steps for the second document as those performed for the first document at block 1202. If process 1200 creates extended discourse trees for more than two documents, process 1200 performs the functions described at block 1202 for multiple documents. Process 1200 may iterate through all pairs of discourse trees in the set of discourse trees, where each discourse tree corresponds to a document. A pair of discourse trees may be represented as follows:

[0111] DT i and DT j ∈DTA At block 1204, the process 1200 includes determining entities and corresponding first basic discourse units from the first discourse tree. Various methods can be used, such as keyword processing (searching for one of a list of predefined keywords in sentences of the first document), using a trained machine learning model, searching internet resources, etc. The discourse navigation application 102 generates the discourse tree DT i and DT j Identify all noun phrases and named entities in

[0112] In one example, the discourse navigation application 102 extracts noun phrases from a discourse tree, and then classifies the noun phrases as (i) entities or (ii) not entities by using a trained machine learning model.

[0113] At block 1205, the process 1200 generates a first base discourse in the second discourse tree. More specifically, the discourse navigation application 102 calculates the overlap and determines the second basic discourse unit that matches the discourse unit. i and DT j Common entity E between i,j The discourse navigation application 102 identifies entities, such as equals, subentities, or parts of i,j The discourse navigation application 102 then establishes relationships between occurrences of entities in E i,j For each occurrence of an entity pair in i,j ) is formed.

[0114] At block 1206, in response to determining a rhetorical connection between the first basic discourse unit and the second basic discourse unit, the process 1200 links the first discourse tree and the second discourse tree via the rhetorical connection, thereby creating an extended discourse tree. More specifically, the discourse navigation application 102 may, for example, i ) and EDU(E j ), construct its DT, and classify the rhetorical relations of each rhetorical link by using the recognized relation labels for each rhetorical link.

[0115] In one aspect, the discourse navigation application 102 combines a first basic discourse unit and a second basic discourse unit into a temporary paragraph, and then determines a rhetorical connection between the first basic discourse unit and the second basic discourse unit in the temporary paragraph by applying a discourse syntactic analysis to the temporary paragraph.

[0116] In a further aspect, in response to not determining a rhetorical connection, the discourse navigation application 102 creates a default rhetorical connection of type elaboration between the first basic discourse unit and the second basic discourse unit, linking the first discourse tree and the second discourse tree.

[0117] In one aspect, the discourse navigation application 102 performs automatic construction and classification of links between text spans across documents, where the following families of approaches can be used: lexical distance, lexical chaining, information extraction, and linguistic template matching. Lexical distance can use cosine similarity across pairs of sentences, while lexical chaining can be more robust, leveraging synonyms and hypernyms.

[0118] An extended discourse tree can form relationships between two or more documents at different levels of granularity. For example, as described with respect to process 1200, relationships can be determined between basic discourse units. Furthermore, an extended discourse tree can represent relationships between words, sentences, paragraphs, sections of documents, or entire documents. As shown, each individual graph is composed of smaller subgraphs for each individual document. Links are shown that represent logical connections between topics within a document.

[0119] Figure 13 also illustrates, according to one aspect, relationships between text units of documents at different levels of granularity. Figure 13 shows discourse trees 1301, 1302, and 1303, each corresponding to a separate document. Figure 13 also illustrates various inter-document links, such as word links 1310 linking words in documents 1302 and 1303, paragraph / sentence links 1311 linking paragraphs or sentences in documents 1301 and 1303, phrase links 1312 linking phrases in documents 1301 and 1303, and cross-document links 1313 linking documents 1301 and 1303. The discourse navigation application 102 uses links 1310-1313 to navigate between documents 1301-1303. You can navigate.

[0120] Using Extended Discourse Trees for Navigation An extended discourse tree such as that produced by process 1200 can be used to navigate a document or other body of text. Extended discourse trees enable a variety of applications, such as autonomous agents, improved search and navigation, and question and answer coordination.

[0121] FIG. 14 illustrates a flowchart of an example process 1400 for navigating between documents using an extended discourse tree, according to one aspect.

[0122] At block 1401, method 1400 includes accessing an expanded discourse tree representing a plurality of documents. As described with respect to process 1200, the expanded discourse tree may include a first discourse tree for a first document and a second discourse tree for a second document, as well as a set of links between the documents that represent rhetorical relationships.

[0123] In one aspect, a document can respond to a particular user question or query. The discourse navigation application 102 can perform a search of a set of documents, a database, or an Internet resource to determine relevant documents. Furthermore, the discourse navigation application 102 can use the question or query as a first document and a document containing an answer to the question or query as a second document.

[0124] At block 1402, the method 1400 includes determining, from the expanded discourse tree, a first basic discourse unit responsive to the query from the user device and a first position corresponding to the first basic discourse unit. Determining the first basic discourse unit may further include matching one or more keywords from the query in the first basic discourse unit. For example, if a threshold number of keywords in the query match a certain basic discourse unit, the basic discourse unit is selected.

[0125] For example, the discourse navigation application 102 receives a user query, e.g., "Atlanta." The discourse navigation application 102 determines a first basic discourse unit that contains the entity "Atlanta." The discourse navigation application 102 then determines an associated location within the first discourse tree. The location can be indicated by various means, such as a node number or an ordered pair that includes a document identifier and a paragraph identifier.

[0126] At block 1403, the method 1400 includes determining a set of navigation options from the expanded discourse tree. The options can include rhetorical relationships between base discourse units within the document, such as a first rhetorical connection between a first base discourse unit and a second base discourse unit in the first discourse tree. The options can also include rhetorical relationships between documents, such as a second rhetorical connection between a first base discourse unit in the first discourse tree and a third base discourse unit in the second discourse tree.

[0127] Continuing with the above example, the discourse navigation application 102 determines that two options are available: one in a first discourse tree that details "Atlanta," e.g., "Atlanta Braves," and one in a second discourse tree that contains more information about "Georgia Tech."

[0128] At block 1404, the method 1400 generates a first rhetorical connection and a second rhetorical connection. Continuing with the example above, the user device 170 presents the "Atlanta Braves" and "Georgia Tech" to the user.

[0129] At block 1405, method 1400 includes (i) presenting a second basic discourse unit to the user device in response to receiving a selection of the first rhetorical connection from the user device, or (ii) presenting a third basic discourse unit to the user device in response to receiving a selection of the second rhetorical connection from the user device.

[0130] Continuing with the above example, the user device 170 receives a selection of "Georgia Tech," and in response, the discourse navigation application 102 provides the user device 170 with a basic discourse unit corresponding to "Georgia Tech," e.g., "Georgia Tech is a research university in Atlanta."

[0131] In one aspect, the discourse navigation application 102 uses the selected results to perform further analysis. For example, based on the selection of "Georgia Tech," the discourse navigation application can search for entities related to "Georgia Tech" in one or more documents, or search for additional documents to analyze and optionally integrate into the expanded discourse tree.

[0132] Applying Extended Discourse Trees to Autonomous Agents Autonomous agents are designed to mimic the intellectual activity of humans maintaining a dialogue. The agents can operate in an iterative manner to provide efficient and effective information to the user. Existing solutions for implementing autonomous agents, including those that use deep learning of word sequences in a dialogue, attempt to construct plausible word sequences to respond to a user's query. In contrast, certain aspects described herein use an extended discourse tree to enable the agent to navigate the user to the appropriate answer as quickly as possible.

[0133] For example, if a user formulates the following query, "Can I pay with one credit card on behalf of another?", the agent attempts to recognize the user's intent and background knowledge about this user to establish an appropriate context. For example, an individual may want to pay with one credit card on behalf of another to avoid late payment fees if cash is unavailable. Instead of providing an answer to this question in the form of a snippet with links to relevant web pages, as major search engines do, certain aspects provide answer topics that the user can choose from. Such topics give the user the opportunity to evaluate, on the one hand, how their request was understood and, on the other hand, what knowledge areas are associated with their question. In our example, topics include "balance transfer," "accessing funds in a checking account," or "canceling your credit card." The user is prompted to select an explanation option, drill down into any of these options, or reject all options and request a new set of topics that the agent can identify.

[0134] Using an expanded discourse tree, the discourse navigation application 102 can start with a root node of the discourse tree, which represents the section of text that most closely matches the user query. The discourse navigation application 102 then builds a set of possible topics by extracting phrases from basic discourse units that are satellites of the root node of the discourse tree. If the user accepts a particular topic, navigation continues along the selected edge of the graph. Otherwise, if the topic does not cover the user's interests, the discourse navigation application 102 can start with an expanded discourse tree. Navigate backward within to another section or another document that matched the original user query.

[0135] Figure 15 illustrates an example of an autonomous agent using an extended discourse tree to answer a user's question, according to one aspect. Figure 15 shows a chat window 1500 containing messages 1501-1506. Messages 1501, 1503, and 1505 are sent by the user device 170, and messages 1502, 1504, and 1506 are sent by the autonomous agent implemented by the discourse navigation application 102.

[0136] As can be seen, the user device 170 initiates a conversation with the agent by sending a message 1501 stating, "I'm broke and short on money." The agent navigates the extended discourse tree, finds topics in a first discourse tree within the extended discourse tree, and determines several topics to respond to message 1501.

[0137] As shown in message 1502, topics include "Out of Your Financial Range," "Bad Decisions Have Consequences," "What I Learned from Being Bankrupt," "Life After Bankruptcy," "Struggling with Estrangement Issues with Various People," and "Limiting Your Current Payments." Each topic is determined by navigating the links in the extended discourse tree. Each topic can be in the first discourse tree or another discourse tree, because the extended discourse tree contains links between and within documents.

[0138] In message 1503, user device 170 selects the "Pay Less" option from the options provided by the agent. The agent then provides the associated paragraph of text to user device 170. The process continues as shown in messages 1504-1506.

[0139] Applying Extended Discourse Trees to Search and Content Discovery On the Web, information is typically represented in a specific section structure in Web pages and documents. Answering questions, forming topics for candidate answers, and attempting to provide an answer based on a user-selected topic are operations that can be represented with the help of a structure that includes a discourse tree of the related text. When a specific portion of text is suggested to a user as an answer, the user may want to drill down to something more specific, raise it to a more general level of knowledge, or move laterally to a topic at the same level. These user intentions to navigate from one portion of text to another can be represented as equal or subordinate discourse relationships between these portions.

[0140] Aspects of the present disclosure improve access times for web-based searches. For example, various aspects can dynamically organize chunks of text from various web pages and documents into a tree, and, in response to a user's selection, the system can navigate to the intended terminal leaf of this tree as quickly as possible. Furthermore, when a user describes their problem in multiple sentences, the autonomous agent attempts to address this problem by finding an answer whose rhetorical structure is in harmony with the structure of the question. In doing so, the agent provides an answer that not only relates to entities from the question, but also is consistent with the logical interrelationships between them.

[0141] Content Discovery In one aspect, the expanded discourse tree is used to facilitate content discovery. In one example, a user device 170 receives a question from a user: "What is faceted searching?" The user wants to understand how faceted searching works. As you are familiar with the concepts, you will want to become familiar with other related concepts.

[0142] In response, the discourse navigation application 102 provides further content exploration or search options. The discourse navigation application 102 determines a set of related documents by forming an extended discourse tree.

[0143] Figure 16 shows an example of an expanded discourse tree, according to one aspect. Figure 16 shows an expanded discourse tree 1600 that includes discourse trees 1602, 1603, and 1604. Each discourse tree 1602-1604 is created from a particular paragraph of text. In this example, the discourse navigation application 102 creates individual discourse trees 1602-1604 from different paragraphs of text. However, units of text of various sizes are possible, such as sentences or multiple paragraphs.

[0144] More specifically, the discourse navigation application 102 may search for the following text related to the topic of faceted search: "Facets correspond to properties of the information elements. They are often derived by analysis of the text of an item using entity extraction techniques or from pre-existing fields in a database such as author, descriptor, language, and format. Thus, existing web-pages, product descriptions, or online collections of articles can be augmented with navigational facets." The discourse tree 1603 is created from the "discourse tree of the item" ("the text of the item"). The discourse tree is derived from the "discourse tree of the item" by analyzing the item's text using feature extraction techniques or from existing fields in the database such as author, descriptor, language, and format. Thus, an online collection of existing web pages, product descriptions, or articles can be extended with navigational facets.

[0145] Additionally, the discourse navigation application 102 may include the following text, also related to the topic of faceted search: "Within the academic community, faceted search has attracted interest primarily among library and information science researchers, but there is a limited interest of computer science researchers specializing in Within academia, faceted search is primarily used in library and information science. Although it has attracted interest among researchers, interest from computer science researchers specializing in information retrieval is limited.

[0146] The discourse navigation application 102 may include the following text related to the topic of entity extraction: "Entity extraction, also known as entity name extraction or named entity recognition, is an information retrieval technique that refers to the The process of identifying and classifying key elements from text into pre-defined categories. (Entity extraction is also known as entity name extraction or named entity recognition.) Discourse tree 1604 is created from the "discourse tree" (also known as "discourse tree recognition"), which is an information retrieval technique that refers to the process of identifying key elements from text and classifying them into predefined categories.

[0147] From the created discourse tree, the discourse navigation application 102 identifies the following additional entities for content exploration: (1) entity extraction, (2) information retrieval, (3) existing fields in the database, and (4) extension with navigation facets. More specifically, the discourse navigation application 102 determines that these entities are related by an elaboration relationship and creates links 1620-1623. The information retrieval represented by node 1611 elaborates on the faceted search represented by node 1610, and thus link 1620 connects nodes 1610 and 1611. The entity extraction represented by node 1613 elaborates on the faceted search represented by node 1612, and thus link 1621 connects nodes 1610 and 1611. 2 and 1613. Information retrieval, represented by node 1615, details entity extraction, node 1614, and thus link 1623 connects nodes 1614 and 1615. Finally, discourse navigation application 102 creates inter-discourse tree link 1622 connecting nodes 1613 and 1615, since discourse tree 1615 details entity extraction.

[0148] The discourse navigation application 102 provides entities to the user device 170. The user device 170 provides the entities to the user, who can follow links to reach a single piece of information or perform a new search and select from multiple search results. For example, starting with a "faceted search," the user device 170 can navigate to information search (e.g., via link 1620 to node 1611), entity extraction (e.g., from node 1612 to node 1613 via link 1621), information search (via link 1622), or further information about the information search (via link 1623 to node 1615).

[0149] Creating additional extended discourse trees The discourse navigation application 102 can build additional expanded discourse trees from existing expanded discourse trees. More specifically, by using the machine learning model 120, the discourse navigation application 102 can create expanded discourse trees based on discourse trees for text in a first domain (e.g., engineering) by using a set of expanded discourse trees for text in a second domain (e.g., law).

[0150] In an exemplary process, the discourse navigation application 102 accesses a first discourse tree representing a first document of a set of documents and a second discourse tree representing a second document from the set of documents.

[0151] Continuing with the example, the discourse navigation application 102 obtains a reference expanded discourse tree from the set of expanded discourse trees by applying the first discourse tree and the second discourse tree to a trained classification model, e.g., machine learning model 120. The set of expanded discourse trees includes multiple expanded discourse trees created by a process such as process 1200. More specifically, the classification model iterates through the set of expanded discourse trees to identify a first candidate discourse tree and a second candidate discourse tree. The classification model identifies the first candidate discourse tree and the second candidate discourse tree as best matches for the first discourse tree and the second discourse tree. The classification model can use different models, such as a classifier or nearest neighbor.

[0152] Continuing with the example, the discourse navigation application 102 determines one or more links between a first reference discourse tree and a second reference discourse tree from the reference extended discourse tree. The links can be determined by using process 1200 (e.g., block 1206). The discourse navigation application 102 then propagates the links to the first discourse tree and the second discourse tree, thereby creating an extended discourse tree. In this manner, the discourse navigation application 102 created an extended discourse tree by identifying an extended discourse tree that includes discourse trees similar to the first and second discourse trees and then generating appropriate inter-discourse tree links.

[0153] Evaluation dataset We used 100 queries and the Clueweb09 cat. B dataset2 (January 2009 to February 2010). 50,220,423 English web pages crawled during the month) We experiment with the TREC datasets from the Web 2009 (queries 1-50) and Web 2010 (queries 51-100) tracks, which collectively include relevance assessments. We chose these datasets because they are widely used in society and can be compared with current state-of-the-art datasets. We use the recommended setting of percentile score <70, which indicates spam3, as proposed by Cormack et al. Spam ranking was used to remove spam.,We consider a subset of this collection consisting of the,top 1000 documents retrieved in response to each query by,the baseline search retrieval model in the tuned settings (described,in Section 4.1.2) using the Indri IR system.

[0154] We created a dataset of Q / A pairs related to car repair recommendations. These pairs were extracted from dialogues as the first and second utterances, so that questions are 7-15 keywords and answers are 3-6 sentences. This resource was obtained to train dialogue support systems, but has also proven useful for evaluating search. This dataset was scraped from (CarPros 2017) and is available at (Github Car Repair Dataset 2017).

[0155] Answer (Webscope 2017) is a set of question-answer pairs covering a wide range of topics. We selected 3,300 questions containing 3-5 sentences from a set of 140,000 user questions. As the answers to most questions are quite detailed, no filtering by sentence length was applied to the answers.

[0156] Our social media dataset primarily includes request-response pairs from posts on Facebook. We also used portions of conversations from LinkedIn.com and vk.com related to employment. In the social domain, writing standards are fairly low. Text cohesion is very limited, and logical structure and relevance are often lacking. The authors formed a training set from their own accounts and from public Facebook accounts available for several years via API (at the time of writing, the Facebook API for retrieving messages is not available). In addition, we used 860 email threads from the Enron dataset (Cohen 2016). Furthermore, we collected data on manual responses to posts from an agent that automatically generates posts on behalf of a human user-host (Galitsky et al. 2014). We collected 4,000 pairs from various social network sources. Formed.

[0157] We created a dataset of financial questions scraped from Fidelity.com. This dataset demonstrates how search relevance improvements occur with reasonable coverage in vertical domains. We compared the efficiency of information access using the proposed chatbot against a major web search engine, such as Google®, for queries for which both systems have relevant answers. In the case of a search engine, a miss is a search result that precedes a search result relevant to a particular user. In the case of a chatbot, a miss is an answer that leads the user to select other options suggested by the chatbot or to request another topic.

[0158] Question topics included personal finances. Twelve users (colleagues of the creator) asked the chatbot 15-20 questions reflecting the user's financial situation, stopping when the user was satisfied with the answers or gave up in dissatisfaction. The same questions were sent to Google, and evaluators had to click on each search result snippet to retrieve a document or webpage and decide whether they were satisfied with it.

[0159] The structure of the comparison of search efficiency between chatbots and search engines is shown in Figure 4. The arrow above shows that all search results (on the left) are used to form a list of topics for explanation. The arrow at the bottom indicates that the answer below was chosen by the chatbot based on two rounds of user feedback and explanation.

[0160] Figure 17 illustrates a comparison between navigation using a search engine and navigation using an autonomous agent enabled with an extended discourse tree, according to one aspect. Figure 17 shows a comparison 1700 that includes a question 1701 posed to a search engine, results 1702-1705 collected in response to the search, and interactions 1701-1706 between a user and an autonomous agent. The arrows indicate how multiple search results on distinct topics converge into a single clarification request that lists automatically extracted topics.

[0161] Rather than finding relevant results by examining all search results (using a search engine on the left), users answer clarification requests generated by the chatbot to drill down to topics of interest to them (on the right). The arrows show how multiple search results for distinct topics converge into a single clarification request that lists automatically extracted topics. The selected topic then navigates the user to a new document or a new section of the same document.

[0162] [Table 3]

[0163] Table 4 shows that chatbot knowledge-seeking sessions last longer than search engines. While this may seem less beneficial to users, businesses prefer users to stay longer on their websites because it increases the likelihood of user acquisition. Spending 7% more time reading chatbot responses is expected to increase users' domain familiarity, especially when these responses conform to the user's preferences. The number of steps in a chatbot search session is one-quarter of those required by a search engine. Traditional methods for measuring search engine performance, such as MAP and NDCG, can also be applied to compare traditional search engines and chatbots in terms of information access efficiency (Sakai 2007). We conclude that using chatbots with augmented discourse tree-driven navigation is an efficient and useful information access method compared to traditional search engines and chatbots that focus on mimicking human intellectual activity.

[0164] 18 is a simplified diagram showing a distributed system 1800 for implementing one of the above aspects. In the illustrated aspect, the distributed system 1800 includes one or more client computing devices 1802, 1804, 1806, and 1808 configured to execute and operate client applications, such as a web browser, a proprietary client (e.g., Oracle Forms), etc., via one or more networks 1810. A server 1812 may be communicatively coupled to the remote client computing devices 1802, 1804, 1806, and 1808 via the network 1810.

[0165] In various aspects, the server 1812 may be adapted to run one or more services or software applications provided by one or more of the system's components. The services or software applications may include non-virtual and virtual environments. The virtual environments may be two-dimensional or three-dimensional (3D) environments. These services may include those used for virtual events, trade shows, simulators, classrooms, purchasing transactions, and business activities, whether as presentations, page-based logical environments, etc. In some aspects, these services may be provided as web-based services or cloud services, or under a Software as a Service (SaaS) model, and may be accessed via client computing devices. The client applications may be provided to users of the client computing devices 1802, 1804, 1806, and / or 1808. Users operating the client computing devices 1802, 1804, 1806, and / or 1808 may then utilize one or more client applications to interact with the server 1812 and utilize the services provided by these components.

[0166] In the illustrated configuration, software components 1818, 1820, and 1822 of distributed system 1800 are shown implemented on server 1812. In other aspects, one or more of the components of distributed system 1800 and / or the services provided by these components may be implemented by one or more of client computing devices 1802, 1804, 1806, and / or 1808. A user operating a client computing device may then utilize one or more client applications to use the services provided by these components. These components may be implemented in hardware, firmware, software, or a combination thereof. It should be understood that a variety of different system configurations are possible that may differ from distributed system 1800. Therefore, the illustrated aspect is an example of a distributed system for implementing the system of the aspect and is not intended to be limiting.

[0167] Client computing devices 1802, 1804, 1806, and / or 1808 may be handheld mobile devices (e.g., iPhone®, mobile phone, iPad®, computing tablet, personal digital assistant (PDA)) or wearable devices (e.g., Google Glass® head-mounted display) that run software such as Microsoft Windows® Mobile® and / or various mobile operating systems such as iOS, Windows Phone, Android, BlackBerry 10, Palm OS, and provide access to the Internet, email, short message service (SMS), BlackBerry®, and other mobile devices. The client computing device may be a general-purpose personal computer, which may, by way of example, be running Microsoft Windows, Apple Macintosh, and / or Linux operating systems. The client computing devices 1802, 1804, 1806, and 1808 may include personal and / or laptop computers running various versions of the UNIX or UNIX-like operating system. The client computing devices may be workstation computers running any of a variety of commercially available UNIX or UNIX-like operating systems, including, but not limited to, various GNU / Linux operating systems such as Google Chrome OS. Alternatively, or additionally, the client computing devices 1802, 1804, 1806, and 1808 may be thin client computers, Internet-enabled gaming systems (e.g., Microsoft Xbox game consoles with or without Kinect gesture input devices), and / or other electronic devices such as personal messaging devices capable of communicating over the network 1810.

[0168] Although the exemplary distributed system 1800 is shown with four client computing devices, any number of client computing devices may be supported. Other devices, such as devices with sensors, may interact with the server 1812.

[0169] Network 1810 in distributed system 1800 may be any type of network familiar to those skilled in the art capable of supporting data communications using any of a variety of commercially available protocols, including, but not limited to, TCP / IP (Transmission Control Protocol / Internet Protocol), SNA (Systems Network Architecture), IPX (Internet Packet Exchange), AppleTalk, etc. By way of example only, network 1810 may be a local area network (LAN), such as one based on Ethernet, token ring, etc. Network 1810 may also be a wide area network and the Internet. Network 1810 may include virtual networks, such as virtual private networks (VPNs). VPN), intranets, extranets, public switched telephone networks (PSTN), infrared networks, wireless networks (e.g., US Telecommunications Networks), Institute of Electrical and Electronics (IEEE) 802.18 The network may include, but is not limited to, networks operating under any of the following protocols: Bluetooth and / or other wireless protocols; and / or any combination thereof; and / or other networks.

[0170] The servers 1812 may be comprised of one or more general-purpose computers, dedicated server computers (including, by way of example, PC (personal computer) servers, UNIX servers, mid-range servers, mainframe computers, rack-mounted servers, etc.), server farms, server clusters, or other suitable configurations and / or combinations. The servers 1812 may include one or more virtual machines running a virtual operating system or other computing architectures involving virtualization. One or more flexible pools of logical storage may be virtualized to maintain virtual storage devices for the servers. Virtual networks may be controlled by the servers 1812 using software-defined networking. In various aspects, the servers 1812 may be adapted to execute one or more services or software applications described in the above disclosure. For example, the servers 1812 may correspond to servers for executing the above-described processes according to aspects of the present disclosure.

[0171] Server 1812 may run an operating system, including any of those listed above, as well as any commercially available server operating system. 2 may run any of a variety of additional server applications and / or middle-tier applications, including a hypertext transport protocol (HTTP) server, a file transfer protocol (FTP) server, a common gateway interface (CGI) server, a JAVA server, a database server, etc. Exemplary database servers are provided by companies such as Oracle, Microsoft, Sybase, It is commercially available from companies such as Sybase and IBM (International Business Machines). This includes, but is not limited to:

[0172] In some implementations, server 1812 may include one or more applications for analyzing and consolidating data feeds and / or event updates received from users of client computing devices 1802, 1804, 1806, and 1808. By way of example, the data feeds and / or event updates may include, but are not limited to, Twitter® feeds, Facebook® updates, or real-time updates received from one or more third-party sources and continuous data streams. Instead, the data feeds may include real-time events related to sensor data applications, financial tickers, network performance measurement tools (e.g., network monitoring and traffic management applications), clickstream analysis tools, automobile traffic monitoring, etc. Server 1812 may also include one or more applications for displaying data feeds and / or real-time events via one or more display devices of client computing devices 1802, 1804, 1806, and 1808.

[0173] Distributed system 1800 may also include one or more databases 1814 and 1816. Databases 1814 and 1816 may reside in a variety of locations. By way of example, one or more of databases 1814 and 1816 may reside on non-transitory storage media local to (and / or resident in) server 1812. Alternatively, databases 1814 and 1816 may be remote from server 1812 and communicate with server 1812 via a network-based or dedicated connection. In one set of aspects, databases 1814 and 1816 may reside on a storage-area network (SAN). Similarly, any necessary files for performing functions attributed to server 1812 may be stored locally and / or remotely on server 1812, as appropriate. In one set of aspects, databases 1814 and 1816 may include relational databases, such as those provided by Oracle Corporation, adapted to store, update, and retrieve data in response to SQL-formatted commands.

[0174] 19 is a simplified block diagram of one or more components of a system environment 1900 in which services provided by one or more components of an aspect of a system can be delivered as cloud services, according to an aspect of the present disclosure. In the aspect shown, the system environment 1900 includes one or more client computing devices 1904, 1906, and 1908 that can be used by users to interact with a cloud infrastructure system 1902 that provides cloud services. The client computing devices can be configured to run client applications, such as a web browser, a proprietary client application (e.g., Oracle Forms), or other applications, that can be used by users of the client computing devices to interact with the cloud infrastructure system 1902 to use services provided by the cloud infrastructure system 1902.

[0175] It should be understood that the cloud infrastructure system 1902 depicted in the figure may have components other than those shown. Moreover, the illustrated aspect is only one example of a cloud infrastructure system that may incorporate aspects of the present invention. In some other aspects, the cloud infrastructure system 1902 may have more or fewer components than those depicted in the figure, may combine two or more components, or may have a different configuration or arrangement of components.

[0176] Client computing devices 1904, 1906, and 1908 may be similar devices to those described above for 2802, 2804, 2806, and 2808.

[0177] Although the exemplary system environment 1900 is shown with three client computing devices, any number of client computing devices may be supported. Other devices, such as devices with sensors, etc., may interact with the cloud infrastructure system 1902.

[0178] Network 1910 may facilitate communication and exchange of data between clients 1904, 1906, and 1908 and cloud infrastructure system 1902. Each network may be any type of network familiar to those skilled in the art capable of supporting data communication using any of a variety of commercially available protocols, including those described above for network 1910.

[0179] Cloud infrastructure system 1902 may comprise one or more computers and / or servers, which may include those described above for server 1712.

[0180] In certain aspects, the services provided by a cloud infrastructure system may include numerous services available on demand to users of the cloud infrastructure system, such as online data storage and backup solutions, web-based email services, hosted office suites and document collaboration services, database processing, and managed technical support services. The services provided by the cloud infrastructure system are dynamically scalable to meet the needs of its users. A specific instantiation of a service provided by the cloud infrastructure system is referred to herein as a "service instance." In general, any service available to users from a cloud service provider's system over a communications network such as the Internet is referred to as a "cloud service." Typically, in a public cloud environment, the servers and systems comprising the cloud service provider's system are distinct from the customer's own on-premises servers and systems. For example, a cloud service provider's system may host an application, and users may order and use the application on demand over a communications network such as the Internet.

[0181] In some examples, services in a computer network cloud infrastructure may include storage, hosted databases, hosted web servers, protected computer network access to software applications, or other services provided to users by cloud vendors or otherwise known in the art. For example, a service may include password-protected access to remote storage on the cloud over the Internet. As another example, a service may include a web-services-based hosted relational database and scripting language middleware engine for private use by networked developers. As another example, the services may include access to an email software application hosted on a cloud vendor's website.

[0182] In certain aspects, cloud infrastructure system 1902 may include a suite of application, middleware, and database service offerings delivered to customers in a self-service, subscription-based, elastically scalable, reliable, highly available, and secure manner. One example of such a cloud infrastructure system is the Oracle Public Cloud offered by the present assignee.

[0183] Large amounts of data, sometimes referred to as big data, can be hosted and / or manipulated by infrastructure systems at multiple levels and at different scales. Such data can include data sets so large and complex that they can be difficult to process using typical database management tools or traditional data processing applications. For example, terabytes of data may be difficult to store, retrieve, and process using personal computers or their rack-based counterparts. Data of this size can be difficult to work with using modern relational database management systems and desktop statistics and visualization packages. They may require massively parallel processing software running thousands of server computers, beyond the architecture of commonly used software tools, to capture, curate, manage, and process the data within acceptable elapsed times.

[0184] Analysts and researchers can store and process extremely large data sets to visualize, detect trends, and / or interact with large amounts of data. Dozens, hundreds, or even thousands of processors linked in parallel can act on such data, thereby displaying it or simulating forces on it or what it represents. These data sets may involve structured data, such as data organized in databases or according to structured models, and / or unstructured data (e.g., emails, images, data blobs (binary large objects), web pages, complex event processing, etc.). By enhancing the ability of a system to relatively quickly converge more (or fewer) computing resources on a target, cloud infrastructure systems can be better utilized to perform tasks on large data sets based on requests from businesses, government agencies, research organizations, private individuals, like-minded groups of individuals or organizations, or other entities.

[0185] In various aspects, cloud infrastructure system 1902 may be adapted to automatically provision, manage, and track customer subscriptions to services provided by cloud infrastructure system 1902. Cloud infrastructure system 1902 may provide cloud services through a variety of deployment models. For example, services may be provided under a public cloud model, where cloud infrastructure system 1902 is owned by an organization that sells cloud services (e.g., owned by Oracle Corporation) and the services are available to the general public or various industry enterprises. As another example, services may be provided under a private cloud model, where cloud infrastructure system 1902 is operated solely for a single organization and may provide services to one or more entities within that organization. Cloud services may also be provided under a community cloud model, where cloud infrastructure system 1902 and the services provided by cloud infrastructure system 1902 are shared by several organizations within an associated community. Cloud services may also be provided through a combination of two or more different models. Cloud services may be provided under a hybrid cloud model.

[0186] In some aspects, the services provided by the cloud infrastructure system 1902 are software as a service (Software as a Service). SaaS Category, Platform as a Service Other categories of services, including PaaS (PaaS), Infrastructure as a Service (IaaS), or hybrid services A subscription order may include one or more services offered under a category. A customer may order one or more services offered by cloud infrastructure system 1902 through a subscription order. Cloud infrastructure system 1902 then performs processing to provide the services in the customer's subscription order.

[0187] In some aspects, services provided by the cloud infrastructure system 1902 may include, but are not limited to, application services, platform services, and infrastructure services. In some examples, application services may be provided by the cloud infrastructure system via a SaaS platform. The SaaS platform may be configured to provide cloud services that fall into the SaaS category. For example, the SaaS platform may provide functionality for building and delivering a suite of on-demand applications on an integrated development and deployment platform. The SaaS platform may manage and control the underlying software and infrastructure for providing the SaaS services. By utilizing the services provided by the SaaS platform, customers can utilize applications that run on the cloud infrastructure system. Customers can obtain application services without the need for customers to purchase separate licenses and support. A variety of different SaaS services may be offered. Examples include, but are not limited to, services that provide solutions for sales performance management, enterprise integration, and business flexibility for large organizations.

[0188] In some aspects, platform services may be provided by a cloud infrastructure system via a PaaS platform. The PaaS platform may be configured to provide cloud services that fall into the PaaS category. Examples of platform services include, but are not limited to, services that enable organizations (e.g., Oracle) to integrate existing applications on a shared, common architecture and the ability to build new applications that leverage shared services provided by the platform. The PaaS platform may manage and control the underlying software and infrastructure to provide the PaaS services. Customers can obtain PaaS services provided by the cloud infrastructure system without the need for the customers to purchase separate licenses and support. Examples of platform services include, but are not limited to, Oracle Java Cloud Service (JCS) and Oracle Database Cloud Service (DBCS).

[0189] By using the services provided by the PaaS platform, customers can use programming languages ​​and tools supported by the cloud infrastructure system and also control the deployed services. In some aspects, the platform services provided by the cloud infrastructure system include database cloud services, middleware cloud services, and the like. These cloud services may include cloud services (e.g., Oracle Fusion middleware services) and Java cloud services. In one aspect, database cloud services may support a shared service deployment model that enables organizations to pool database resources and provide database-as-a-service services to customers in the form of a database cloud. Middleware cloud services may provide customers with a platform for developing and deploying various business applications in a cloud infrastructure system, and Java cloud services may provide customers with a platform for deploying Java applications in a cloud infrastructure system.

[0190] A variety of different infrastructure services may be provided by the IaaS platform in a cloud infrastructure system. The infrastructure services facilitate the management and control of basic computing resources such as storage, network, and other underlying computing resources for customers who use the services provided by the SaaS and PaaS platforms.

[0191] Additionally, in certain aspects, cloud infrastructure system 1902 may include infrastructure resources 1930 for providing resources used to provide various services to customers of the cloud infrastructure system. In one aspect, infrastructure resources 1930 may include a pre-integrated, optimized combination of hardware, such as servers, storage, and networking resources, for running the services provided by the PaaS and SaaS platforms.

[0192] In some aspects, resources in cloud infrastructure system 1902 may be shared by multiple users and dynamically reallocated on a demand basis. Resources may also be allocated to users in different time periods. For example, cloud infrastructure system 1930 may maximize resource utilization by allowing a first set of users in a first time period to utilize resources of the cloud infrastructure system for a specified time period and allowing reallocation of the same resources to another set of users located in a different time period.

[0193] In certain aspects, several internal shared services 1932 may be provided that are shared by various components or modules of, and services provided by, cloud infrastructure system 1902. These internal shared services may include, but are not limited to, security and identity services, integration services, enterprise repository services, enterprise manager services, virus scanning and whitelisting services, high availability, backup and recovery services, services to enable cloud support, email services, notification services, file transfer services, etc.

[0194] In certain aspects, cloud infrastructure system 1902 may provide comprehensive management of cloud services (e.g., SaaS services, PaaS services, and IaaS services) in the cloud infrastructure system. In one aspect, cloud management functionality may include functionality for provisioning, managing, tracking, etc., customer subscriptions received by cloud infrastructure system 1902.

[0195] In one aspect, as shown in the figure, the cloud management functionality includes an order management module 1919, an order orchestration module 1922, an order provisioning module 1930, and a cloud management function 1940. The functions and features of the system may be provided by one or more modules, such as a processing module 1924, an order management and monitoring module 1926, and an identity management module 1928. These modules may include or be provided using one or more computers and / or servers, which may be general-purpose computers, dedicated server computers, server farms, server clusters, or other suitable configurations and / or combinations.

[0196] In example operation 1934, a customer using a client device, such as client device 1904, 1906, or 1908, may interact with cloud infrastructure system 1902 by requesting one or more services offered by cloud infrastructure system 1902 and placing an order for a subscription to one or more services provided by cloud infrastructure system 1902. In particular aspects, the customer may access a cloud user interface (UI), i.e., cloud UI 1919, cloud UI 1914, and / or cloud UI 1916, and place a subscription order via these UIs. Order information received by cloud infrastructure system 1902 in response to the customer placing an order may include information identifying the customer and the one or more services offered by cloud infrastructure system 1902 to which the customer intends to subscribe.

[0197] After an order is placed by a customer, the order information is received via the cloud UI 1919, 1914 and / or 1916.

[0198] In operation 1936, the order is stored in an order database 1919. The order database 1919 may be one of several databases operated by the cloud infrastructure system 1919 and in cooperation with other system elements.

[0199] In operation 1938, the order information is forwarded to the order management module 1919. In some examples, the order management module 1919 may be configured to perform billing and accounting functions related to the order, such as confirming the order and booking the order upon confirmation.

[0200] At operation 1940, information about the order is communicated to order orchestration module 1922. The order orchestration module 1922 may utilize the order information to orchestrate the provisioning of services and resources for the order placed by the customer. In some examples, the order orchestration module 1922 may orchestrate the provisioning of resources to support the subscribed service using the services of the order provisioning module 1924.

[0201] In certain aspects, the order orchestration module 1922 enables management of business processes associated with each order and applies business logic to determine whether the order should proceed to provisioning. In operation 1942, upon receiving an order for a new subscription, the order orchestration module 1922 sends a request to the order provisioning module 1924 to allocate resources and configure those resources needed to fulfill the subscription order. The order provisioning module 1924 enables allocation of resources for services ordered by a customer. The order provisioning module 1924 processes the services ordered by the cloud infrastructure system 1900. The order orchestration module 1922 provides a level of abstraction between the cloud services provided by the cloud service provider and the physical implementation layer used to provision the resources to provide the requested services. Thus, the order orchestration module 1922 can be decoupled from implementation details such as whether services and resources are actually provisioned on the fly or pre-provisioned and only allocated / assigned when requested.

[0202] In operation 1944 , once the services and resources are provisioned, notification of the provided services may be sent by the order provisioning module 1924 of the cloud infrastructure system 1902 to the customer on the client device 1904 , 1906 and / or 1908 .

[0203] At operation 1946, the customer's subscription order may be managed and tracked by the order management and monitoring module 1926. In some examples, the order management and monitoring module 1926 may be configured to collect usage statistics for the services in the subscription order, such as the amount of storage used, the amount of data transferred, the number of users, and the amount of system up time and system down time.

[0204] In certain aspects, cloud infrastructure system 1900 may include identity management module 1928. Identity management module 1928 may be configured to provide identity services, such as access management and authorization services, in cloud infrastructure system 1900. In some aspects, identity management module 1928 may control information about customers who wish to use services provided by cloud infrastructure system 1902. Such information may include information authenticating the identities of such customers and information describing which actions those customers are authorized to perform on various system resources (e.g., files, directories, applications, communication ports, memory segments, etc.). Identity management module 1928 may also include management of descriptive information about each customer and how and by whom this descriptive information may be accessed and modified.

[0205] 20 illustrates an exemplary computer system 2000 in which various aspects of the present invention may be implemented. System 2000 may be used to implement any of the computer systems described above. As shown, computer system 2000 includes a processing unit 2004 that communicates with several peripheral subsystems via a bus subsystem 2002. These peripheral subsystems may include a processing acceleration unit 2006, an I / O subsystem 2008, a storage subsystem 2018, and a communications subsystem 2024. Storage subsystem 2018 includes a tangible computer-readable storage medium 2022 and a system memory 2010.

[0206] Bus subsystem 2002 provides a mechanism for allowing the various components and subsystems of computer system 2000 to communicate with each other as intended. While bus subsystem 2002 is shown schematically as a single bus, alternative implementations of the bus subsystem may utilize multiple buses. Bus subsystem 2002 may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures. For example, such architectures may include the Industry Standard Architecture (ISA) bus, Micro Channel architecture, which may be implemented as a mezzanine bus manufactured to the IEEE P2086.1 standard. Examples of such buses include the Multi-Channel Architecture (MCA) bus, the Enhanced ISA (EISA) bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0207] Processing unit 2004, which may be implemented as one or more integrated circuits (e.g., conventional microprocessors or microcontrollers), controls the operation of computer system 2000. Processing unit 2004 may include one or more processors. These processors may include single-core or multi-core processors. In certain aspects, processing unit 2004 may be implemented as one or more independent processing units 2032 and / or 2034, each with a single-core or multi-core processor included therein. In other aspects, processing unit 2004 may be implemented as a quad-core processing unit formed by incorporating two dual-core processors on a single chip.

[0208] In various aspects, the processing unit 2004 may execute various programs in response to program code and may maintain multiple programs or processes running simultaneously. At any given time, some or all of the program code to be executed may reside in the processor 2004 and / or the memory subsystem 2018. Through suitable programming, the processor 2004 may provide the various functions described above. The computer system 2000 may also additionally include a processing acceleration unit 2006, which may include a digital signal processor (DSP), special purpose processor, etc.

[0209] The I / O subsystem 2008 may include user interface input devices and user interface output devices. User interface input devices may include keyboards, pointing devices such as mice or trackballs, touchpads or touchscreens, which may incorporate displays, scroll wheels, click wheels, dials, buttons, switches, keypads, audio input devices, as well as voice command recognition systems, microphones, and other types of input devices. User interface input devices may include motion detection and / or gesture recognizers, such as a Microsoft Kinect® motion sensor, which allows a user to control and interact with an input device, such as a Microsoft Xbox® 360 game controller, through a natural user interface using gestures and spoken commands. User interface input devices may also include eye gesture recognizers, such as a Google Glass® blink detector, which detects eye movements from a user (e.g., "blinks" while taking pictures and / or making menu selections) and translates the eye gestures as input to the input device. The user interface input devices may also include a voice recognition sensing device that allows a user to interact with a voice recognition system (e.g., the Siri® navigator) via voice commands.

[0210] User interface input devices may also include, but are not limited to, three-dimensional (3D) mice, joysticks or pointing sticks, gamepads and graphics tablets, and audio / visual devices such as speakers, digital cameras, digital camcorders, portable media players, webcams, image scanners, fingerprint scanners, barcode readers, 3D scanners, 3D printers, laser range finders, and eye gaze detection devices. User interface input devices may also include medical imaging input devices, such as, for example, computed tomography, magnetic resonance imaging, positional emission tomography, and medical ultrasound equipment. User interface input devices may also include, for example, medical imaging input devices, such as, for example, computed tomography, magnetic resonance imaging, positional emission tomography, and medical ultrasound equipment. For example, this may include audio input devices such as MIDI keyboards, digital musical instruments, and the like.

[0211] User interface output devices may include display subsystems, indicator lights, or non-visual displays such as audio output devices. The display subsystem may be a flat panel display, such as one using a cathode ray tube (CRT), liquid crystal display (LCD), or plasma display, a projection device, a touch screen, or the like. In general, use of the term "output device" is intended to include all possible types of devices and mechanisms for outputting information from computer system 2000 to a user or another computer. For example, user interface output devices may include, but are not limited to, various display devices that visually convey text, graphics, and audio / video information, such as monitors, printers, speakers, headphones, automobile navigation systems, plotters, voice output devices, and modems.

[0212] The computer system 2000 may include a storage subsystem 2018 that includes software elements shown as being currently located within system memory 2010. The system memory 2010 may store program instructions that are loadable and executable on the processing unit 2004, and data generated during the execution of these programs.

[0213] Depending on the configuration and type of the computer system 2000, the system memory 2010 may be volatile (such as random access memory (RAM)) and / or non-volatile (such as read-only memory (ROM), flash memory, etc.). RAM typically houses data and / or program modules that are immediately accessible to the processing unit 2004, and / or data and / or program modules that are currently being operated on and executed by the processing unit 2004. In some implementations, the system memory 2010 may include multiple different types of memory, such as static random access memory (SRAM) or dynamic random access memory (DRAM). In some implementations, a basic input / output system (BIOS) that includes basic routines to assist in transferring information between elements within the computer system 2000, such as during startup, may typically be stored in ROM As an example and without limitation, the system memory 2010 may include client applications, web browsers, middleware applications, relational database management systems (RDBMS), etc. Also shown are possible application programs 2012, program data 2014, and operating system 2016. By way of example, operating system 2016 may include various versions of the Microsoft Windows®, Apple Macintosh®, and / or Linux operating systems, various commercially available UNIX® or UNIX-like operating systems (including, but not limited to, various GNU / Linux operating systems, Google Chrome® OS, etc.), and / or mobile operating systems such as iOS, Windows® Phone, Android® OS, BlackBerry® 10 OS, and Palm® OS operating systems.

[0214] Additionally, the storage subsystem 2018 may provide a tangible, computer-readable storage medium for storing the basic programming and data structures that provide the functionality of certain aspects. Software (programs, code modules, instructions) that, when executed by a processor, provide the functionality described above may be stored in the storage subsystem 2018. These software modules or instructions may be executed by the processing unit 2004. The storage subsystem 2018 may also provide a repository for storing data used in accordance with the present invention.

[0215] The storage subsystem 2000 may also include a computer-readable storage medium reader 2020 further connectable to a computer-readable storage medium 2022. Together and optionally in combination with the system memory 2010, the computer-readable storage medium 2022 may comprehensively represent remote, local, fixed, and / or removable storage devices, as well as storage media for temporarily and / or permanently containing, storing, transmitting, and retrieving computer-readable information.

[0216] The computer-readable storage medium 2022 containing the code or portions of code may include any suitable medium known or used in the art, including storage and communication media such as, but not limited to, volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storing and / or transmitting information. This may include tangible, transitory computer-readable storage media such as RAM, ROM, electronically erasable programmable ROM (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disk (DVD), or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage, or other magnetic storage, or other tangible computer-readable media. It may also include intangible, transitory computer-readable media such as data signals, data transmissions, or any other medium usable to transmit the desired information and accessible by the computing system 2000.

[0217] By way of example, the computer-readable storage medium 2022 may include a hard disk drive that reads from or writes to non-removable, non-volatile magnetic media, a magnetic disk drive that reads from or writes to removable, non-volatile magnetic disks, and an optical disk drive that reads from or writes to removable, non-volatile optical disks such as CD-ROMs, DVDs, and Blu-ray disks or other optical media. The computer-readable storage medium 2022 may include, but is not limited to, Zip drives, flash memory cards, universal serial bus (USB) flash drives, secure digital (SD) cards, DVD disks, digital video tapes, and the like. The computer-readable storage medium 2022 may also include flash memory-based SSDs, enterprise flash drives, solid-state drives (SSDs) based on non-volatile memory such as solid-state ROMs, solid-state These may include SSDs based on volatile memory such as RAM, dynamic RAM, static RAM, DRAM-based SSDs, magnetoresistive RAM (MRAM) SSDs, and hybrid SSDs that use a combination of DRAM and flash memory-based SSDs. The disk drives and their associated computer-readable media may provide computer-readable instructions, data structures, program modules, and other data to the computer system 2000.

[0218] The communications subsystem 2024 provides an interface with other computer systems and networks. The communications subsystem 2024 acts as an interface for receiving data from other systems and transmitting data from the computer system 2000 to other systems. For example, the communications subsystem 2024 may enable the computer system 2000 to connect to one or more devices via the Internet. In some aspects, the communications subsystem 2024 may be used with other technologies (e.g., cellular technologies such as 3G, 4G, or EDGE (enhanced data rates for global evolution) or advanced digital technologies (e.g., cellular technologies such as 3G, 4G, or EDGE (enhanced data rates for global evolution)). a radio frequency (RF) transceiver component for accessing wireless voice and / or data networks (using WiF data network technology); i (IEEE 802.28 family of standards or other mobile communication technologies or any combination thereof), global positioning system (GPS) receiver control In some aspects, the communications subsystem 2024 may provide a wired network connection (e.g., Ethernet) in addition to or instead of a wireless interface.

[0219] Additionally, in some aspects, the communications subsystem 2024 may receive incoming communications in the form of structured and / or unstructured data feeds 2026, event streams 2028, event updates 2020, etc., on behalf of one or more users who may be using the computer system 2000.

[0220] As an example, the communications subsystem 2024 may provide Twitter feeds, Facebook updates, Rich Site Summary (RSS) feeds, and other similar services. The data feeds 2026 may be configured to receive data feeds, such as web feeds, in real time from users of social media networks and / or other communication services, and / or to receive real-time updates from one or more third-party sources.

[0221] Additionally, the communications subsystem 2024 may be configured to receive data in the form of a continuous data stream. The data may include an event stream 2028 of real-time events and / or event updates 2020 that may be continuous or may be essentially unbounded with no clear ends. Examples of applications that generate continuous data may include, for example, sensor data applications, financial tickers, network performance measurement tools (e.g., network monitoring and traffic management applications), clickstream analysis tools, automobile traffic monitoring, etc.

[0222] The communications subsystem 2024 may also be configured to output structured and / or unstructured data feeds 2026, event streams 2028, event updates 2020, etc. to one or more databases that may communicate with one or more streaming data source computers coupled to the computer system 2000.

[0223] The computer system 2000 may be one of a variety of types, including a handheld mobile device (e.g., an iPhone® mobile phone, an iPad® computing tablet, a PDA), a wearable device (e.g., a Google Glass® head-mounted display), a PC, a workstation, a mainframe, a kiosk, a server rack, or other data processing system.

[0224] Due to the ever-changing nature of computers and networks, the description of computer system 2000 shown in the figure is intended only as a specific example. Many other configurations are possible having more or fewer components than the system shown in FIG. 20. For example, customized hardware may be used and / or particular elements may be implemented in hardware, firmware, software (including applets), or a combination. Additionally, connections to other computing devices, such as network input / output devices, may be utilized. Based on the disclosure and teachings provided herein, one of ordinary skill in the art will recognize other means and / or methods for implementing various aspects.

[0225] In the foregoing specification, aspects of the invention have been described with reference to specific aspects thereof. Those skilled in the art will recognize that the present invention is not limited in this regard. Various features and aspects of the above-described invention may be used individually or together. Moreover, aspects may be utilized in a number of environments and applications beyond those described herein without departing from the broader spirit and scope of the specification. Accordingly, the specification and drawings are to be regarded as illustrative rather than restrictive.

Claims

1. 1. A computer-implemented method for navigating text using an extended discourse tree, comprising: accessing expanded discourse trees representing a plurality of documents, the expanded discourse trees including a first discourse tree for a first document and a second discourse tree for a second document, the method further comprising: determining from the expanded discourse tree (i) a first basic discourse unit that includes an entity included in a query from a user device, and (ii) each location of a link representing each of one or more rhetorical connections associated with the first basic discourse unit; performing discourse parsing to determine from the expanded discourse tree a set of navigation options including (i) a first rhetorical connection, the first rhetorical connection being one of the one or more rhetorical connections between the first basic discourse unit and a second basic discourse unit of the first discourse tree, and (ii) a second rhetorical connection, the second rhetorical connection being one of the one or more rhetorical connections between the first basic discourse unit and a third basic discourse unit of the second discourse tree; presenting a first entity included in the second basic discourse unit and a second entity included in the third basic discourse unit to a user device; A computer-implemented method comprising: (i) presenting the second basic discourse unit to the user device in response to receiving a selection of the first entity from the user device; or (ii) presenting the third basic discourse unit to the user device in response to receiving a selection of the second entity from the user device.

2. 2. The computer-implemented method of claim 1, wherein each of the first discourse tree and the second discourse tree includes a plurality of nodes, each non-terminal node of the plurality of nodes representing a rhetorical connection, and each terminal node of the nodes of the respective discourse tree being associated with a basic discourse unit.

3. 3. The computer-implemented method of claim 1, further comprising, in response to receiving an additional query from the user device, determining additional basic discourse units responsive to the additional query and presenting the additional basic discourse units to the user device.

4. 4. The computer-implemented method of claim 1, wherein determining the first basic discourse unit further comprises identifying the first basic discourse unit in the extended discourse tree and matching one or more keywords from the query in the first basic discourse unit.

5. Determining the first basic discourse unit further includes: generating a first parse tree for the query; generating an additional parse tree for each of the one or more basic discourse units; 5. The computer-implemented method of claim 1, further comprising: in response to determining that one of the additional parse trees includes the first parse tree, selecting a basic discourse unit corresponding to the additional parse tree as the first basic discourse unit.

6. 6. The computer-implemented method of claim 1, wherein each of the first and second rhetorical connections is (i) elaboration, (ii) enablement, (iii) conditional, (iv) contrast, or (v) attribution.

7. 7. A computer-implemented method according to claim 1, wherein the first document comprises a first plurality of basic discourse units, the first discourse tree representing a first rhetorical connection between at least two of the first plurality of basic discourse units, and the second document comprises a second plurality of basic discourse units, and the second discourse tree representing a second rhetorical connection between at least two of the second plurality of basic discourse units.

8. 8. The computer-implemented method of claim 7, wherein the expanded discourse tree includes the links representing rhetorical connections between common entities in the first discourse tree and the second discourse tree.

9. 1. A system comprising: a non-transitory computer-readable medium storing computer-executable program instructions; a processing unit communicatively coupled to the non-transitory computer-readable medium for executing the computer-executable program instructions, wherein executing the computer-executable program instructions configures the processing unit to perform operations, the operations including: accessing expanded discourse trees representing a plurality of documents, the expanded discourse trees including a first discourse tree for a first document and a second discourse tree for a second document, the operations further comprising: determining from the expanded discourse tree (i) a first basic discourse unit that includes an entity included in a query from a user device, and (ii) each location of a link representing each of one or more rhetorical connections associated with the first basic discourse unit; performing discourse parsing to determine from the expanded discourse tree a set of navigation options including (i) a first rhetorical connection, the first rhetorical connection being one of the one or more rhetorical connections between the first basic discourse unit and a second basic discourse unit of the first discourse tree, and (ii) a second rhetorical connection, the second rhetorical connection being one of the one or more rhetorical connections between the first basic discourse unit and a third basic discourse unit of the second discourse tree; presenting a first entity included in the second basic discourse unit and a second entity included in the third basic discourse unit to a user device; (i) presenting the second basic discourse unit to the user device in response to receiving a selection of the first entity from the user device, or (ii) presenting the third basic discourse unit to the user device in response to receiving a selection of the second entity from the user device.

10. 10. The system of claim 9, wherein the operations further include, in response to receiving an additional query from the user device, determining additional basic discourse units responsive to the additional query and presenting the additional basic discourse units to the user device.

11. The system of claim 9 or 10, wherein determining the first basic discourse unit further comprises matching one or more keywords from the query in the first basic discourse unit.

12. Determining the first basic discourse unit further includes: generating a first parse tree for the query; generating an additional parse tree for each of the one or more basic discourse units; 12. The system of claim 9, further comprising: in response to determining that one of the additional parse trees includes the first parse tree, selecting a basic discourse unit corresponding to the additional parse tree as the first basic discourse unit.

13. The system of any of claims 9 to 12, wherein each of the first and second rhetorical connections is (i) an elaboration, (ii) an enablement, (iii) a condition, (iv) a contrast, or (v) an attribution.

14. The system of any one of claims 9 to 13, wherein the first document includes a first plurality of basic discourse units, the first discourse tree representing a first rhetorical connection between at least two of the first plurality of basic discourse units, and the second document includes a second plurality of basic discourse units, and the second discourse tree representing a second rhetorical connection between at least two of the second plurality of basic discourse units.

15. A program for causing a computer to execute the method according to any one of claims 1 to 8.

Images