Method for guessing next wanted instruction of user based on relaticsearch and current instruction request semantic analysis result

Abstract

The invention discloses a method for guessing an instruction which a user wants to use next based on an elasticsearch and a current instruction request semantic analysis result. The method comprises the following steps: determining a next round of instruction field; preparing a next round of instruction; instruction information of the round is extracted; and updating data in real time. According to the invention, based on the Elasticsearch and the user request instruction, the next round of prompt can be quickly assembled, the prompt rule and the user-defined rule can be updated in real time, and the difficulty of using the semantic instruction by the user can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical fields of natural semantic processing and artificial intelligence, and more specifically, to a method for guessing the next instruction that a user wants to use based on elasticsearch and the semantic analysis result of the current instruction request. Background Art

[0002] In current smart TVs, many users only shout some simple instructions. Although the next-round help prompts on the TV are configured and pushed to the terminal by operators, this requires a large amount of manpower. The next-round help prompts may not be what the user needs and have little relation to the semantics of what the user shouts this time. The user experience is very poor and it is time-consuming and laborious for application operators. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for guessing the next instruction that a user wants to use based on elasticsearch and the semantic analysis result of the current instruction request, in order to solve the technical problems existing in the background art.

[0004] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0005] A method for guessing the next instruction that a user wants to use based on elasticsearch and the semantic analysis result of the current instruction request, comprising:

[0006] Determining the next-round instruction field;

[0007] Preparing for the next-round instruction;

[0008] Extracting the information of the current-round instruction;

[0009] Updating data in real time.

[0010] In some embodiments, the determining the next-round instruction field includes: based on the user's historical viewing records, obtaining the user behavior data within a cycle behavior, obtaining the instructions with a higher frequency of the user; finding the corresponding users according to the instructions, and determining the field of the next-round instruction through the instructions shouted by the users subsequently; finally, selecting the top six fields with the highest appearance of BAIKE in the next round and storing them in redis as the next-round fields.

[0011] In some embodiments, the preparing for the next-round instruction includes: counting the same-type instructions in different fields, extracting the entities in the instructions, and performing semantic parsing on the user instruction request to parse out the corresponding entities; as new fields appear, adding new instruction rules to elasticsearch so that the next-round prompts of semantics can be updated in time.

[0012] In some embodiments, the user instruction request needs to be semantically parsed to parse out the corresponding entity. After parsing out the corresponding entity, the instruction is regularized as follows: replace "film" after parsing out the corresponding entity and assemble the next-round instruction; after extracting the instruction rules according to the domain intention, store them in Elasticsearch.

[0013] In some embodiments, the extraction of the current-round instruction information includes: extracting relevant entities according to the current-round instruction and parsing out the corresponding domain intention; requesting media asset data according to the domain intention and the relevant corresponding entity.

[0014] In some embodiments, the film title, actors, directors, characters, types, play sources, categories, introductions, release times, and regions of the corresponding entity are returned in the media asset data; the entities in the media asset data are extracted.

[0015] In some embodiments, the real-time update of data includes: when a new domain appears, real-time update the corresponding domain array in Redis; when a new statement appears, real-time update the rule instruction data in Elasticsearch.

[0016] The beneficial effects of the present invention compared with the prior art are as follows:

[0017] Through the analysis of the user's viewing history data, the corresponding rules of the present invention are stored in Elasticsearch, and combined with the user's current round, corresponding prompts are given to the user to continue using the TV, which can improve the user's TV experience and reduce the operation burden at the same time.

[0018] The design of this invention is based on Elasticsearch and user request instructions, realizing the rapid assembly of the next-round prompt and the real-time update of the prompt rules and custom rules, and reducing the difficulty for users to use semantic instructions. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a step diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions in the preferred embodiments of the present application will be described in more detail below with reference to the accompanying drawings in the preferred embodiments of the present application. In the drawings, the same or similar reference numerals denote the same or similar components or components with the same or similar functions throughout. The described embodiments are some, but not all, of the embodiments of the present application. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present application and should not be construed as limiting the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the scope of protection of the present application.

[0021] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0022] In the description of the present application, it should be noted that, unless otherwise clearly defined and limited, the terms "installation", "connection", and "connection" should be understood in a broad sense. For example, it can be a fixed connection or an indirect connection through an intermediate medium, and it can be the communication inside two components or the interaction relationship between two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific situations.

[0023] In the description of the present application, it should be understood that the orientation or positional relationship indicated by the terms "upper", "lower", "front", "rear", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship in the accompanying drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or component referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present application.

[0024] In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or display that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or displays.

[0025] The following will be combined with Figure 1 , and a method for conjecturing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request in the embodiments of the present application will be described in detail. It should be noted that the following embodiments are only used to explain the present application and do not constitute a limitation to the present application.

[0026] Embodiment 1:

[0027] As Figure 1 shown, a method for conjecturing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request includes the following steps:

[0028] (1) Determination of the next-round instruction field;

[0029] The determination of the next-round field depends on the user's historical viewing records. First, obtain the user behavior data within the cycle behavior and obtain the instructions with a higher frequency for the user. Based on these instructions, find the corresponding users, and determine the field of the next-round instruction through the instructions that the user continues to shout next, as shown in Table 1 below:

[0030] Table 1

[0031] This round This round of instructions Next round Next round of instructions BAIKE Who is Liu XX VIEDO Watch Liu XX's movies BAIKE Who is Liu XX MUSIC Listen to Liu XX's songs BAIKE Who is Liu XX TRAILER On which channel is Liu XX's movie broadcast BAIKE Who is Liu XX BAIKE Liu XX's height BAIKE Who is Liu XX NEWS News related to Liu XX BAIKE Who is Liu XX TV Brightness increase

[0032] Finally, select the top six fields with the highest appearance in the next round of BAIKE and store them as the next-round fields in redis. For example: BAIKE: [VIEDO, MUSIC, TRAILER, BAIKE, NEWS, TV].

[0033] (2) Next-round instruction preparation

[0034] Count the same-type instructions for different fields as shown in Table 2 below:

[0035] Table 2

[0036]

[0037]

[0038] Extract the entities in the instruction

[0039] The user instruction request needs to be semantically parsed to parse out the corresponding entities. Regularize the instruction as follows:

[0040] I want to watch [film]

[0041] The [film] in the instruction rule corresponds to the title of the movie. Different users call different movies. After parsing out the corresponding entities, film can be replaced and the next-round instruction can be assembled.

[0042] After extracting these instruction rules according to the domain intention, they need to be stored in elasticsearch. The storage is as follows:

[0043]

[0044] As new fields appear or new statements appear, elasticsearch can easily add new instruction rules to update the next-round hint of semantics in a timely manner.

[0045] (3) Extraction of this-round instruction information

[0046] According to this-round instruction, extract the relevant entities and parse out the corresponding domain intention;

[0047] According to the domain intention and the relevant corresponding entities, request the media asset data. The media asset data returns the title, actors, directors, characters, types, playback sources, categories, introductions, release times, regions, etc. of the corresponding entities; Extract the entities in the media asset data, such as:

[0048] {

[0049] "actor":"Sun|Liu|Ma|Fang|Huang|Gao|Jiang|Sun",

[0050] "year":"2015",

[0051] "desc":"During the Warring States Period, Mi Yue was the favorite little princess of King Wei of Chu. However, her status plummeted when King Wei of Chu went to war. Her mother Xiang was expelled from the palace by Queen Wei of Chu. Many years later, she returned to the palace to avenge her life. Mi Yue and Huang Xie, the prince of Chu, were childhood sweethearts and truly loved each other. In order to elope with Huang Xie smoothly, Mi Yue volunteered to marry the legitimate princess Mi Shu to the State of Qin. On the way to Qin, Mi Yue and Mi Shu supported each other. In the middle of the way, the Chu State's ceremonial carriage was robbed by the army led by Yiqu King Zhai Li. Huang Xie fell into a valley in order to save Mi Yue and his life and death were unknown. Mi Yue was disheartened. , in order to find out the mastermind behind the scenes, she accompanied Mi Shu into the Qin Palace. Mi Shu became the Queen of Qin. Mi Yue was framed by Lady Wei and her half brother Wei Ran was kidnapped. She had no choice but to ask the King of Qin to become his favorite concubine. The original sisterly love gradually split after Mi Yue gave birth to her son Ying Ji. The sons fought for the throne, and King Qin Ying Si died with regret. Mi Yue and her son were exiled to the distant Yan State. Unexpectedly, King Qin Wu Ying Dang died while lifting the tripod, and Qin was in chaos. Mi Yue returned to Qin with the help of Yiqu’s military force and quelled the civil strife in Qin. Mi Yue’s son Ying Ji ascended the throne and became king, known in history as King Zhaoxiang of Qin. Mi Yue became the first queen mother in history, known in history as Queen Dowager Xuan of Qin. ",

[0052] "film":"Legend of Miyue",

[0053] "type":"TV series",

[0054] "tag":"Ancient Costumes|Dragon Boat Festival Special|TV Series"

[0055] }

[0056] (4) Update data in real time

[0057] When a new field appears, the corresponding field array in redis is updated in real time.

[0058] When new statements appear, the rule instruction data in elasticsearch is updated in real time.

[0059] At present, smart TVs are very popular, and help prompts are also widely used in TV products. In order to better enable users to find the movies they want when using TV. The inventor assembles the next round of instructions based on the user's historical viewing records and the user's current command results, and gives the user a prompt on how to shout the next round of commands. Figure 1 The specific principles are explained as follows:

[0060] 1. Obtain the result of the current instruction: After the user requests the central control semantics, the central control calls the semantic component to parse the corresponding domain, intent, and corresponding entities, and then requests media data according to the domain intent and entities. Extract the corresponding entities in the media data, such as entity data of movie title, actors, directors, characters, types, etc., and merge them into the entity data parsed by semantics.

[0061] 2. Obtain the domain for the next round: Use the domain parsed in the previous step as the key to query the array of domains that can be traversed in the next round in redis.

[0062] 3. Obtain the rule instructions: According to the domain data obtained in the previous step, randomly select 3 domains and query the data corresponding to the domains in elasticsearch.

[0063] 4. Assemble the instructions for the next round: Combine the entities in the first step and the rule instructions queried in the third step, and loop through to replace the identifiers in the rules. After the replacement is completed, remove duplicates from the assembled instruction array, and at the same time exclude instructions that are the same as the current instruction. Finally, randomly select 3 and return them to the terminal, such as:

[0064]

[0065] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for guessing the next command that a user wants to use based on elasticsearch and the semantic analysis results of the current command request, characterized in that: include: The next round of directive areas is determined; Prepare for the next round of instructions; Extraction of command information for this round; Update data in real time.

2. According to claim 1, a method for guessing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request is characterized in that: The next round of instruction domain is determined; including: based on the user's historical viewing record, obtaining user behavior data within the periodic behavior, obtaining instructions with higher user frequency; finding the corresponding user according to the instruction, and determining the domain of the next round of instructions through the instructions shouted by the user subsequently; finally selecting the top six domains with the highest appearance in the next round of BAIKE and storing them in redis as the next round of domains.

3. According to claim 1, a method for guessing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request is characterized in that: The preparation of the next round of instructions includes: counting the same type of instructions in different fields, extracting the entities in the instructions, and performing semantic analysis on user instruction requests to parse out the corresponding entities; as new fields emerge, elasticsearch adds new instruction rules so that the next round of semantic prompts can be updated in time.

4. According to claim 3, a method for guessing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request is characterized in that: The user instruction request needs to be semantically parsed to parse out the corresponding entities, wherein the instructions are regularized, such as: after parsing out the corresponding entities, replace the film and assemble the next round of instructions; after extracting the instruction rules according to the domain intent, store them in elasticsearch.

5. According to claim 1, a method for guessing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request is characterized in that: The extraction of the instruction information of this round includes: extracting relevant entities and parsing corresponding domain intentions according to the instructions of this round; and requesting media resource data according to the domain intentions and the corresponding entities.

6. The method for guessing the next command that the user wants to use based on elasticsearch and the semantic analysis result of the current command request according to claim 5 is characterized in that: The film title, actor, director, role, type, source, category, introduction, release time, and region of the corresponding entity are returned in the media asset data; the entity in the media asset data is extracted.

7. The method for guessing the next instruction that the user wants to use based on elasticsearch and the semantic analysis result of the current instruction request according to claim 1 is characterized in that: The real-time data updating includes: when a new field appears, updating the corresponding field array in redis in real time; when a new statement appears, updating the rule instruction data in elasticsearch in real time.

Images