Live room filter selection method and device, electronic equipment and readable medium

Abstract

Embodiments of the present application provide a live room filter selection method and device, electronic equipment and readable medium, the method comprises: before the host live broadcast, by acquiring the live room information of the live room, then according to the live room information, generate the word array matched with the live room, then the word array is matched with the preset filter vocabulary library, obtain the corresponding filter word, the target filter corresponding to the filter word is used as the filter used after the live broadcast, so as to utilize the live room information to extract features, obtain the word array matched with the live room, then the word array is matched with the filter vocabulary library before the live room broadcast, obtain the filter word, to obtain the corresponding target filter, ensure that the filter effect is consistent with the theme of the live room, the target filter is used as the default filter when the live room is broadcast, avoid the host to switch the filter in the process of live broadcast, simplify the process of adding filter.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of live broadcast, in particular to a live broadcast room filter selection method, a live broadcast room filter selection device, an electronic device and a computer readable medium. BACKGROUND

[0002] With the rapid development of image processing technology, people in the process of taking a selfie or recording a video through a camera, real-time beautifying adjustment processing of the character in the picture has become the norm. Beautifying refers to changing the skin color of the character in the picture through an algorithm, such as eliminating acne marks on the face, adjusting the color tone of the entire picture or a part of the picture through filter effects, etc. Different filter effects can bring different beautifying effects. In the live broadcast industry, in order to present a better picture to the audience and bring a good viewing experience to the audience, the host usually uses filter effects to match the current live broadcast scene. However, in the process of selecting filter effects by the host, on the one hand, the host needs to manually set the corresponding filter effects, which can easily lead to excessive time consumption of the host in the live broadcast preparation stage, and on the other hand, if there is a network delay during the live broadcast process, the filter effects cannot be added in time, which can easily affect the live broadcast of the host and bring poor live broadcast experience to the host. SUMMARY

[0003] The embodiments of the present application provide a live broadcast room filter selection method, device, electronic device and computer readable storage medium to solve or partially solve the problem that the filter addition process is complicated and prone to failure due to network influence in the process of adding filter effects to the live broadcast room in the related art.

[0004] The embodiments of the present application disclose a live broadcast room filter selection method, comprising:

[0005] obtaining live broadcast room information of a live broadcast room;

[0006] generating a word array matched with the live broadcast room according to the live broadcast room information;

[0007] matching the word array with a preset filter vocabulary library to obtain a filter word;

[0008] using a target filter corresponding to the filter word as a filter used after the live broadcast room starts broadcasting.

[0009] Optionally, the step of generating a word array matched with the live broadcast room according to the live broadcast room information comprises:

[0010] performing word segmentation on the live broadcast room information to obtain a plurality of first words;

[0011] Respectively, the first word is tagged with a part of speech, and a word array matched with the live room is generated.

[0012] Optionally, the live room information at least includes introduction information and a room name of the live room, and the live room information is segmented to obtain a plurality of first words, which includes:

[0013] The introduction information and / or the room name is segmented to obtain a plurality of first words.

[0014] Optionally, the respective first word is tagged with a part of speech, and a word array matched with the live room is generated, which includes:

[0015] The part of speech of each first word is determined.

[0016] Each of the words and the corresponding part of speech of the word is composed into a first key-value pair.

[0017] The TF-IDF word frequency feature and the annotation sequence feature of each first word are obtained, and the word frequency feature and the annotation sequence feature are used to calculate the weight coefficient of each first word.

[0018] The first key-value pair is sorted using the weight coefficient to generate a word array matched with the live room.

[0019] Optionally, the word array is matched with a preset filter vocabulary library to obtain a filter word, which includes:

[0020] Extracting a target word of a target part of speech in the word array;

[0021] Obtaining the original text corresponding to the target word;

[0022] If the target word is determined to be available using the original text, the target word is matched with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0023] Optionally, the target part of speech includes adjectives and nouns, and the target word of the target part of speech in the word array includes:

[0024] Traverse the part of speech corresponding to the first word in the word array;

[0025] The first word with the part of speech of adjective or noun is taken as the target word.

[0026] Optionally, if the target word is determined to be available using the original text, the target word is matched with a preset filter vocabulary library to obtain a filter word corresponding to the target word, which includes:

[0027] if the target word corresponds to an odd number of negative words in the original text, determining that the target word is unusable;

[0028] if the target word corresponds to an even number of negative words in the original text, determining that the target word is usable, and matching the target word with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0029] Optionally, the matching the target word with the preset filter vocabulary library to obtain the filter word corresponding to the target word comprises:

[0030] searching, from the preset filter vocabulary library, a filter word same as, synonymous with, or near-synonymous with the target word as the filter word.

[0031] Optionally, the method further comprises:

[0032] if there is no filter word matching the target word in the filter vocabulary library, selecting a preset filter as the filter used after the live room goes on air.

[0033] Optionally, the filter vocabulary library is generated in the following manner:

[0034] obtaining a filter of the live room and at least one second word corresponding to the filter;

[0035] forming a second key-value pair with the filter and the second word corresponding to the filter;

[0036] generating the filter vocabulary library by using each second key-value pair.

[0037] Optionally, the generating the filter vocabulary library by using each second key-value pair comprises:

[0038] obtaining a use frequency of each filter;

[0039] sorting the second key-value pairs according to the use frequency to generate the filter vocabulary library.

[0040] The embodiment of the application further discloses a selection device for filters of a live room, comprising:

[0041] a live room information obtaining module configured to obtain live room information of a live room;

[0042] a word array generating module configured to generate a word array matched with the live room according to the live room information;

[0043] a filter word obtaining module configured to match the word array with a preset filter vocabulary library to obtain a filter word.

[0044] a filter determination module, configured to determine a target filter corresponding to the filter word as a filter used after the live room goes on air.

[0045] Optionally, the word array generation module comprises:

[0046] a first word obtaining sub-module, configured to tokenize the live room information to obtain a plurality of first words;

[0047] a word array generation sub-module, configured to respectively perform part-of-speech tagging on each of the first words to generate a word array matched with the live room.

[0048] Optionally, the live room information at least comprises introduction information and a room name of the live room, and the first word obtaining sub-module is specifically configured to:

[0049] tokenize the introduction information and / or the room name to obtain a plurality of first words.

[0050] Optionally, the word array generation sub-module is specifically configured to:

[0051] determine the part-of-speech of each of the first words;

[0052] compose a first key-value pair with each of the words and the part-of-speech corresponding to the word;

[0053] obtain a TF-IDF word frequency feature and a labeled sequence feature of each of the first words, and calculate a weight coefficient of each of the first words by using the word frequency feature and the labeled sequence feature;

[0054] sort the first key-value pair by using the weight coefficient to generate a word array matched with the live room.

[0055] Optionally, the filter word obtaining module comprises:

[0056] a target word extracting sub-module, configured to extract a target word of a target part-of-speech from the word array;

[0057] an original text obtaining sub-module, configured to obtain an original text corresponding to the target word;

[0058] a filter word obtaining sub-module, configured to, if it is determined by using the original text that the target word is available, match the target word with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0059] Optionally, the target part-of-speech comprises an adjective and a noun, and the target word extracting sub-module is specifically configured to:

[0060] traversing a part-of-speech corresponding to the first word in the word array;

[0061] taking the first word with a part-of-speech of adjective or noun as a target word.

[0062] Optionally, the filter word obtaining submodule is specifically configured to:

[0063] if the target word has a corresponding odd number of negative words in the original text, determining that the target word is unusable;

[0064] if the target word has a corresponding even number of negative words in the original text, determining that the target word is usable, and matching the target word with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0065] Optionally, the filter word obtaining submodule is specifically configured to:

[0066] searching, from the preset filter vocabulary library, a filter word that is the same as, synonymous with, or near-synonymous with the target word as the filter word.

[0067] Optionally, the method further comprises:

[0068] a preset filter selecting module configured to, if there is no filter word matching the target word in the filter vocabulary library, select a preset filter as a filter to be used after the live room goes on air.

[0069] Optionally, the filter vocabulary library is generated by:

[0070] a word obtaining module configured to obtain a filter of the live room and at least one second word corresponding to the filter;

[0071] a key-value pair forming module configured to form a second key-value pair of the filter and the second word corresponding to the filter;

[0072] a filter vocabulary library generating module configured to generate the filter vocabulary library by using each of the second key-value pairs.

[0073] Optionally, the filter vocabulary library generating module is specifically configured to:

[0074] obtain a frequency of use of each of the filters;

[0075] sort the second key-value pairs according to the frequency of use to generate the filter vocabulary library.

[0076] An embodiment of the present invention further discloses an electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0077] The memory is used to store computer programs;

[0078] The processor is configured to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0079] The embodiments of the present invention further disclose one or more computer-readable media having instructions stored thereon. When executed by one or more processors, the processors are enabled to perform the method according to the embodiments of the present invention.

[0080] The embodiments of the present invention include the following advantages:

[0081] In an embodiment of the present invention, before the host starts live broadcasting, the live broadcast client can first obtain the live broadcast room information of the live broadcast room, and then generate a word array matching the live broadcast room according to the live broadcast room information, and then match the word array with a preset filter vocabulary library to obtain corresponding filter words. The live broadcast client can use the target filter corresponding to the filter word as the filter used after the live broadcast starts, thereby using the live broadcast room information to perform feature extraction and obtain a word array matching the live broadcast room. Then, before the live broadcast room starts, the word array is matched with the filter vocabulary library to obtain the filter word to obtain the corresponding target filter, thereby ensuring the consistency of the filter effect with the live broadcast room theme. When the live broadcast room starts, the target filter is used as the default filter, which avoids the host from switching filters during the live broadcast, simplifies the process of adding filters, and improves the success rate of adding filters. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 This is a flowchart of the steps of a method for selecting a live broadcast room filter provided in an embodiment of the present invention;

[0083] Figure 2 This is a structural block diagram of a device for selecting a live broadcast filter provided in an embodiment of the present invention;

[0084] Figure 3 is a block diagram of an electronic device provided in an embodiment of the present invention;

[0085] Figure 4 is a schematic diagram of a computer-readable medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0086] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0087] As an example, in the live broadcast industry, in order to present a better picture to the audience and bring a good viewing experience to the audience, the host usually uses a filter to match the current live broadcast scene during the live broadcast. However, in the process of selecting a filter, on the one hand, the host needs to manually set the corresponding filter, which is likely to cause the host to spend too much time in the live broadcast preparation stage, and on the other hand, if there is a network delay during the live broadcast process, the filter cannot be added in time, which is likely to affect the host to live broadcast and bring a poor live broadcast experience to the host.

[0088] To this end, one of the core points of the embodiments of the present application is to determine a filter according to the live room information of a live room before the live room starts broadcasting, so that the filter can be directly used after the live room starts broadcasting, thereby avoiding the host to set the filter during the live broadcast process. Specifically, the word array can be determined according to the live room information before the live room starts broadcasting, and the word array is matched with the filter vocabulary library to obtain filter words to obtain the corresponding target filter, so as to ensure the consistency of the filter effect and the theme of the live room. The target filter is used as the default filter when the live room starts broadcasting, which avoids the host to switch the filter during the live broadcast process, simplifies the process of adding the filter, and improves the success rate of adding the filter.

[0089] Specifically, referring to Figure 1 , a step flowchart of a method for selecting a filter of a live room according to an embodiment of the present application is shown, which can specifically include the following steps:

[0090] Step 101, obtaining live room information of a live room;

[0091] In the embodiments of the present application, a host user (hereinafter referred to as a host) can run a corresponding live broadcast client in a host terminal to realize live broadcast of corresponding content, such as game live broadcast, outdoor live broadcast, food live broadcast, house source live broadcast, and car source live broadcast. For a viewer user, he can watch the real-time live broadcast of the host through a watching client and interact with the host or other viewers in the live room through a barrage, a virtual gift, and the like.

[0092] Optionally, the terminal can be a mobile phone, a tablet computer, a portable computer, a desktop computer, a personal digital assistant, a wearable device (such as glasses, a watch, etc.), and the like. The live broadcast client can be an application program running in the terminal, such as a live broadcast application program. The host can create a corresponding live room in the live broadcast application program to live broadcast corresponding content in the live room.

[0093] For the live room, the host can set the related live room information, introduce the live room, so as to attract users, or facilitate users to find the live room they want to watch according to the live room information. Among them, the live room information can include introduction information and room name, etc. The introduction information can be the information added by the host before the broadcast according to the content of the live broadcast. Through the introduction information, the user can understand the theme and content of the live broadcast. The role of the room name is similar to the introduction information, which will not be described here. Through the live room information, the audience can be simply and clearly informed of the live content corresponding to the live room. For example, the introduction information can be "outdoor live, hope to get everyone's attention", which can be used to tell the audience that the live room is live outdoors; the introduction information can be "one person playing four rows on the island", which can be used to tell the audience that the live room is live game.

[0094] Step 102, generating a word array matched with the live room according to the live room information;

[0095] In the embodiment of the application, after the live client obtains the live room information set by the host in the live room, the live room information can be segmented, a plurality of first words are obtained, and then each first word is tagged with a part of speech, to generate a word array matched with the live room. The purpose of part-of-speech tagging is to mark each first word with a separate label, which can represent the usage and syntactic role, such as noun, verb, adjective, preposition, etc.

[0096] Specifically, if the host sets the introduction information of the live room, before the live room starts, the live client can segment the introduction information to obtain a plurality of first words corresponding to the introduction information, or segment the room name to obtain a plurality of first words corresponding to the room name, or segment both or other information used to introduce the live room, to obtain a plurality of first words corresponding to the live room. Optionally, the live room information about the live room can be information in the form of text, which is not limited by the application.

[0097] After obtaining a plurality of first words corresponding to the live room information, the part of speech of each first word can be analyzed and tagged, such as verb, adjective, noun, preposition, etc. After determining the part of speech of each first word, the first key-value pair can be formed by each first word and its corresponding part of speech. The TF-IDF can be used to analyze each first word corresponding to the live room information, to obtain the TF-IDF word frequency feature and the labeled sequence feature of each first word. The word frequency feature and the labeled sequence feature are used to calculate the weight coefficient of each first word. Then, the first key-value pair can be sorted by using the weight coefficient, to generate a word array matched with the live room.

[0098] It should be noted that the TF-IDF: calculation formula is TF*IDF, TF (Term Frequency, word frequency), generally the number of occurrences of a word in a sentence divided by the total number of words in the sentence; IDF (Inverse Document Frequency), its size is inversely proportional to the common degree of a word, indicating the "importance" weight of each word, the most common word is given the smallest weight, the more common word is given smaller weight, and the less common word is given larger weight.

[0099] In an example, assuming that the anchor sets the introduction information as "sunny and not hot, outdoor live broadcast, please support" according to the content to be live broadcast before the live broadcast, after the introduction information is segmented, the words "sunny", "bright", "not", "hot", "outdoor", "live broadcast", "please", "more" and "support" can be obtained, and then the part-of-speech tagging can be performed to obtain the following table 1:

[0100] Word Part of speech Sunlight Noun Bright Adjective Not Verb Hot Adjective Outdoor Adverb Live Verb Please Verb Much Adverb Support Verb

[0101] Table 1

[0102] Among them, each first word and part-of-speech can be combined in the form of key-value pair, with word as key and part-of-speech as value, such as [sunny: noun], [bright: adjective] and the like, and then the weight coefficient obtained by TF-IDF analysis can be sorted to obtain the word array shown in the above table 1, so that the live room information is segmented and part-of-speech tagged, so that the matching filter is selected according to the part-of-speech of each first word, and then the filter matched with the live room information is used after the live room starts, to ensure the live effect of the anchor.

[0103] Step 103, matching the word array with the preset filter vocabulary library to obtain filter words;

[0104] In the embodiment of the application, for the filter effects that can be used in the live room, the corresponding second words can be set for each filter effect in advance to generate a filter vocabulary library for the filter effect, so that after the word array is determined according to the live room information, it can be matched with the filter vocabulary library to determine whether there is a corresponding filter word, if there is, the target filter corresponding to the filter word can be used as the filter used after the live room starts; if not, the default filter effect can be directly used, and the anchor can select the filter according to the actual demand, so as to ensure the consistency of the filter effect and the theme of the live room, and the target filter is used as the default filter when the live room starts, avoiding the anchor to switch the filter during the live broadcast, simplifying the process of adding the filter, and improving the success rate of adding the filter.

[0105] Optionally, the filter vocabulary library can be generated by obtaining a filter of a live broadcast room and at least one filter vocabulary corresponding to the filter, then grouping the filter and the filter vocabulary corresponding to the filter into a second key-value pair, and then generating the filter vocabulary library by using each second key-value pair. Specifically, for a live broadcast room, a developer has configured corresponding filter effects in the client during the development of the live broadcast client, such as different filter effects of "shining", "dark tone", "natural", and "cold". For these filter effects, the second words (including synonyms, synonyms, and the same words, etc.) corresponding to them can be obtained by word forest, machine learning, etc., and then the corresponding key-value pair can be established with the filter effect as the key and the second word as the value, and all key-value pairs are combined to generate the filter vocabulary library. For example, the filter style can be sorted by frequency of use, and the synonyms of the style are sorted in alphabetical order. The synonyms are added to the array synonymArray in alphabetical order, and the key-value pair [style:synonymArray] is added to the dictionary styleLibrary in the order of filter style usage frequency from high to low to form the filter vocabulary library to improve search efficiency, as shown in Table 2 below:

[0106] Filter effect Second word Shimmering "Shimmering", "Shining", "Glimmering", "Glimmer", "Star" Dark tone "Dark tone", "Dark", "Deep" Natural "Natural", "Natural", "Natural" Cold "Chilly", "Chilly", "Chilly", "Chilly", "Chilly" … …

[0107] Table 2

[0108] The representation of the key-value pair can be:

[0109] "shining": ["shining", "shining", "shining", "light", "star point"];

[0110] "dark tone": ["dark tone", "dark color", "deep"];

[0111] "natural": ["natural", "natural", "natural"];

[0112] "cold": ["cold", "cold", "cold", "cold", "cold"];

[0113] Among them, one filter effect can correspond to several different words. By matching the filter effect and the corresponding word, and generating the filter vocabulary library according to a certain sorting rule, the efficiency of extracting filter words can be effectively improved, and the host can use the corresponding filter effect when broadcasting, improving the host's live broadcast experience.

[0114] Before the live broadcast room starts, the live broadcast client can extract the target word with the target part of speech from the word array according to the live broadcast room information set by the host, and obtain the original text corresponding to the target word. Then, if the original text is used to determine that the target word is available, the target word is matched with the filter vocabulary library to obtain the filter word corresponding to the target word. The original text can be the introduction information itself, or the room name of the live broadcast room, etc.

[0115] In a specific implementation, the second word corresponding to the filter effect can be an adjective or a noun, etc. Before the vocabulary matching, it is necessary to traverse the part of speech of all words in the word array, and the first word with the part of speech of adjective or noun is taken as the target word. The word with a specific part of speech is extracted as the word matched with the filter vocabulary library, so that the feature information in the live broadcast room information is used for filter selection. On the one hand, it can effectively ensure that the obtained filter matches the theme of the live broadcast room, and ensure the live broadcast effect of the host. On the other hand, different live broadcast room information can correspond to different filters, so that the filter effect of the live broadcast room is more personalized, and the diversity of the use of the live broadcast room filter is enriched. At the same time, since there can be negative words, if the negative words are combined with adjectives or nouns, the target word extracted does not express the original meaning of the adjective or noun, but its opposite meaning, for example, "hot" and "not hot" are words with opposite meanings. Therefore, before matching the target word with the second word in the filter vocabulary library, it is also necessary to determine whether the target word is available. Specifically, if the target word has a corresponding odd number of negative words in the original text, the target word is determined to be unavailable; if the target word has a corresponding even number of negative words in the original text, the target word is determined to be available, and the target word is matched with the preset filter vocabulary library to obtain the filter word corresponding to the target word. Therefore, by determining whether the extracted word is available, the obtained filter can be further matched with the theme of the live broadcast room, and the live broadcast effect of the host is ensured.

[0116] After the target word is determined, the filter vocabulary library can be searched for a filter word that is the same as, synonymous with, or similar to the target word. If there is no filter word in the filter vocabulary library that matches the target word, a preset filter is selected as the filter used after the live broadcast room starts. The preset filter can be the last filter used by the host in the last live broadcast process, the default filter set by the live broadcast room, or the filter most frequently used by the host in the live broadcast process, and the present application does not limit this.

[0117] In an example, after obtaining the word array abstractArray corresponding to the live room information, elements in the abstractArray can be retrieved, for the i-th element abstractArray[i], if the part of speech of abstractArray[i] is an adjective or a noun, the original text of the synopsis is retrieved, if there are odd number of negative words before abstractArray[i], the word is considered unavailable, if there are even number of negative words before abstractArray[i], the word is considered available, for example, the word "hot" in "the sunshine is not hot" is unavailable, the word "hot" in "the sunshine is not not hot" is available, and the like.

[0118] If the word abstractArray[i] is available, starting from the most commonly used filter style in the filter word library, a binary search is performed in the synonymArray of the style to search for abstractArray[i], if found, the filter style is returned, if the synonymArray of the filter style does not contain the word, the next filter style is searched, if the corresponding word is not found, it is determined that there is none, and a preset filter is used as the filter used after the live room starts, through the binary search, it can be quickly determined whether there is a filter word corresponding to the target word in the filter word library, thereby effectively ensuring that the filter used after the live room starts is determined before the host starts, and the live experience of the host is improved.

[0119] It should be noted that the embodiments of the present application include but are not limited to the above examples, and it can be understood that related modes can also be set by those skilled in the art under the guidance of the idea of the embodiments of the present application, and the present application does not limit this.

[0120] Step 104, the target filter corresponding to the filter word is used as the filter used after the live room starts.

[0121] In a specific implementation, if there is a filter word matching the live room information, the target filter corresponding to the filter word is obtained, and the target filter is set as the default filter of the live room, which is used after the host starts; if there is no filter word matching the live room information, a preset filter is directly used as the default filter of the live room, which is used after the host starts, thereby before the live room starts, the word array is matched with the filter word library to obtain the filter word to obtain the corresponding target filter, ensuring the consistency of the filter effect and the theme of the live room, the target filter is used as the default filter when the live room starts, avoiding the host to switch the filter during the live process, simplifying the process of adding the filter, and the process of determining the live room filter before the host starts is less affected by the network, improving the success rate of adding the filter.

[0122] In the embodiment of the present application, before the host starts live broadcast, the live broadcast client can first acquire the live broadcast room information of the live broadcast room, then generate a word array matched with the live broadcast room according to the live broadcast room information, and then can match the word array with the preset filter vocabulary library to obtain the corresponding filter word. The live broadcast client can use the target filter corresponding to the filter word as the filter used after the live broadcast room starts live broadcast, so as to extract features by using the live broadcast room information to obtain the word array matched with the live broadcast room, then match the word array with the filter vocabulary library before the live broadcast room starts live broadcast to obtain the filter word, so as to obtain the corresponding target filter, ensure the consistency of the filter effect and the theme of the live broadcast room, use the target filter as the default filter when the live broadcast room starts live broadcast, avoid the host switching the filter during live broadcast, simplify the process of adding the filter, and improve the success rate of adding the filter.

[0123] It should be noted that, for the method embodiments, in order to simply describe, they are all described as a series of action combinations, but those skilled in the art should know that the embodiments of the present application are not limited to the action sequence described, because according to the embodiments of the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions involved are not necessarily necessary for the embodiments of the present application.

[0124] Referring to Figure 2 , a structural block diagram of a live broadcast room filter selection device provided in an embodiment of the present application is shown, which can specifically include the following modules:

[0125] A live broadcast room information acquisition module 201 is configured to acquire live broadcast room information of a live broadcast room.

[0126] A word array generation module 202 is configured to generate a word array matched with the live broadcast room according to the live broadcast room information.

[0127] A filter word obtaining module 203 is configured to match the word array with a preset filter vocabulary library to obtain a filter word.

[0128] A filter determination module 204 is configured to use a target filter corresponding to the filter word as a filter used after the live broadcast room starts live broadcast.

[0129] In an optional embodiment of the present application, the word array generation module 202 includes:

[0130] A first word obtaining sub-module is configured to perform word segmentation on the live broadcast room information to obtain a plurality of first words.

[0131] The word array generation submodule is configured to perform part-of-speech tagging on each first word to generate a word array matched with the live room.

[0132] In an optional embodiment of the present application, the live room information at least includes introduction information and a room name of the live room, and the first word obtaining submodule is specifically configured to:

[0133] perform word segmentation on the introduction information and / or the room name to obtain a plurality of first words.

[0134] In an optional embodiment of the present application, the word array generation submodule is specifically configured to:

[0135] determine the part-of-speech of each first word;

[0136] compose a first key-value pair with each word and the part-of-speech corresponding to the word;

[0137] obtain a TF-IDF word frequency feature and a label sequence feature of each first word, and calculate a weight coefficient of each first word by using the word frequency feature and the label sequence feature;

[0138] sort the first key-value pair by using the weight coefficient to generate a word array matched with the live room.

[0139] In an optional embodiment of the present application, the filter word obtaining module 203 includes:

[0140] a target word extraction submodule configured to extract a target word of a target part-of-speech from the word array;

[0141] a raw text obtaining submodule configured to obtain a raw text corresponding to the target word;

[0142] a filter word obtaining submodule configured to, if it is determined that the target word is available by using the raw text, match the target word with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0143] In an optional embodiment of the present application, the target part-of-speech includes an adjective and a noun, and the target word extraction submodule is specifically configured to:

[0144] traverse the part-of-speech of each first word in the word array;

[0145] extract a first word with an adjective or noun as a target word.

[0146] In an optional embodiment of the present application, the filter word obtaining submodule is specifically configured to:

[0147] if the target word corresponds to an odd number of negative words in the original text, determining that the target word is unusable;

[0148] if the target word corresponds to an even number of negative words in the original text, determining that the target word is usable, and matching the target word with a preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0149] In an optional embodiment of the present application, the filter word obtaining submodule is specifically configured to:

[0150] searching, from the preset filter vocabulary library, a filter word same as, synonymous with, or near-synonymous with the target word as the filter word.

[0151] In an optional embodiment of the present application, the present application further comprises:

[0152] a preset filter selecting module configured to select a preset filter as the filter used after the live broadcast room is started if no filter word matching the target word exists in the filter vocabulary library.

[0153] In an optional embodiment of the present application, the filter vocabulary library is generated by the following modules:

[0154] a word obtaining module configured to obtain a filter of the live broadcast room and at least one second word corresponding to the filter;

[0155] a key-value pair forming module configured to form a second key-value pair of the filter and the second word corresponding to the filter;

[0156] a filter vocabulary library generating module configured to generate the filter vocabulary library by using each second key-value pair.

[0157] In an optional embodiment of the present application, the filter vocabulary library generating module is specifically configured to:

[0158] obtain a use frequency of each filter;

[0159] sort the second key-value pairs according to the use frequency to generate the filter vocabulary library.

[0160] For the device embodiment, it is basically similar to the method embodiment, so the description is relatively simple, and the related parts refer to the part of the method embodiment.

[0161] In addition, the embodiments of the present application also provide an electronic device, such as a mobile phone, a tablet computer, a personal computer, a wearable device, or the like. Figure 3As shown, the server comprises a processor 301, a communication interface 302, a memory 303 and a communication bus 304, wherein the processor 301, the communication interface 302 and the memory 303 complete mutual communication through the communication bus 304,

[0162] The memory 303 is used for storing a computer program.

[0163] The processor 301 is used for executing the program stored in the memory 303 to realize the following steps:

[0164] Obtaining live room information of a live room;

[0165] Generating a word array matched with the live room according to the live room information;

[0166] Matching the word array with a preset filter vocabulary library to obtain a filter word;

[0167] Taking a target filter corresponding to the filter word as a filter used after the live room is started.

[0168] In an optional embodiment of the present application, the step of generating a word array matched with the live room according to the live room information comprises:

[0169] Segmenting the live room information to obtain a plurality of first words;

[0170] Respectively marking the parts of speech of each first word to generate a word array matched with the live room.

[0171] In an optional embodiment of the present application, the live room information at least comprises introduction information and a room name of the live room, and the step of segmenting the live room information to obtain a plurality of first words comprises:

[0172] Segmenting the introduction information and / or the room name to obtain a plurality of first words.

[0173] In an optional embodiment of the present application, the step of respectively marking the parts of speech of each first word to generate a word array matched with the live room comprises:

[0174] Determining the parts of speech of each first word;

[0175] Forming a first key-value pair with each word and the part of speech corresponding to the word;

[0176] Obtaining a TF-IDF word frequency feature and a labeled sequence feature of each first word, and using the word frequency feature and the labeled sequence feature to calculate a weight coefficient of each first word;

[0177] The first key-value pair is sorted by using the weight coefficient, and a word array matched with the live broadcast room is generated.

[0178] In an optional embodiment of the present application, the matching of the word array with the preset filter vocabulary library to obtain a filter word comprises:

[0179] extracting a target word of a target part of speech in the word array;

[0180] obtaining original text corresponding to the target word;

[0181] If it is determined that the target word is available by using the original text, the target word is matched with the preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0182] In an optional embodiment of the present application, the target part of speech comprises an adjective and a noun, and the extracting of the target word of the target part of speech in the word array comprises:

[0183] traversing a part of speech corresponding to the first word in the word array;

[0184] the first word with an adjective or a noun as the target word.

[0185] In an optional embodiment of the present application, if it is determined that the target word is available by using the original text, the target word is matched with the preset filter vocabulary library to obtain a filter word corresponding to the target word, which comprises:

[0186] If the target word has a corresponding odd number of negative words in the original text, it is determined that the target word is not available;

[0187] If the target word has a corresponding even number of negative words in the original text, it is determined that the target word is available, and the target word is matched with the preset filter vocabulary library to obtain a filter word corresponding to the target word.

[0188] In an optional embodiment of the present application, the matching of the target word with the preset filter vocabulary library to obtain a filter word corresponding to the target word comprises:

[0189] searching, from the preset filter vocabulary library, a filter word same as, synonymous with, or near-synonymous with the target word as the filter word.

[0190] In an optional embodiment of the present application, the method further comprises:

[0191] If there is no filter vocabulary matching the target word in the filter vocabulary library, a preset filter is selected as the filter used after the live room goes on air.

[0192] In an optional embodiment of the present application, the filter vocabulary library is generated in the following manner:

[0193] Obtaining a filter of the live room and at least one second word corresponding to the filter;

[0194] Grouping the filter and the second word corresponding to the filter into a second key-value pair;

[0195] Generating the filter vocabulary library by using each second key-value pair.

[0196] In an optional embodiment of the present application, the generating the filter vocabulary library by using each second key-value pair comprises:

[0197] Obtaining a use frequency of each filter;

[0198] Sorting the second key-value pairs according to the use frequency to generate the filter vocabulary library.

[0199] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0200] The communication interface is used for communication between the terminal and other devices.

[0201] The memory can include a Random Access Memory (RAM) and can also include a non-volatile memory, such as at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0202] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0203] like Figure 4 As shown, in another embodiment provided by the present invention, a computer-readable storage medium 401 is also provided, in which instructions are stored. When the computer-readable storage medium 401 is run on a computer, the computer executes the method for selecting the live broadcast room filter described in the above embodiment.

[0204] In another embodiment of the present invention, a computer program product including instructions is provided. When the computer program product is run on a computer, the computer executes the method for selecting the live broadcast room filter described in the above embodiment.

[0205] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0206] It is to be noted that, as used in this specification and the appended claims, the singular forms "a," "an" and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a component" can include a combination of two or more components, and the term "an element" can include comparable reference to a plurality of elements. Also, as used in this specification and the appended claims, the term "or" as used in the context of "A / B" or "A / B / C" means any of the possibilities; for example, "A or B" means "A or B or both", and "A or B or C" means "A or B or C or any two or all three of these". Furthermore, to the extent that "comprising", "including", "carrying", "having", "containing", and variants and / or derivatives thereof are used in either the specification or the claims, such terms are intended to be open- ended, that is, to mean including, without being limited to, the stated components. For example, a composition containing ingredients X, Y, and Z would be within the scope of a composition containing ingredients X, Y, and Z or any subset thereof; for example, a composition containing only ingredients X and Y or a composition containing only ingredients X and Z. Where the disclosure includes "and / or", such as in "A and / or B" and "A and B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B. Where the disclosure includes "one", such as in "one of A and B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B. Where the disclosure includes "at least one", such as in "at least one of A and B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B. Where the disclosure includes "one or more", such as in "one or more of A and B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B. Where the disclosure includes "at least one of A or B", such as in "at least one of A or B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B. Where the disclosure includes "one, and only one of A or B", such as in "one, and only one of A or B", such descriptions mean A alone, B alone, or A and B, including embodiments that are a combination of A and B and embodiments that are neither A nor B.

[0207] Each of the various embodiments in the specification are described in a related manner, and the same or similar parts among the various embodiments can be referred to each other. Each of the various embodiments focuses on the difference from other embodiments. In particular, for the system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

[0208] The above only describes the preferred embodiments of the present application, and is

Claims

1. A live room filter selection method, characterized in that, The application relates to a method for generating a filter word corresponding to a live broadcast room. The method comprises the following steps: acquiring live broadcast room information of a live broadcast room, wherein the live broadcast room information comprises introduction information and a room name, and the introduction information is information added according to live broadcast content before starting broadcasting; generating a word array matched with the live broadcast room according to the live broadcast room information, wherein the word array is generated by performing part-of-speech tagging on first words obtained by dividing the live broadcast room information into words, and the part-of-speech of the first words comprises verbs, adjectives, nouns and prepositions; matching the word array with a preset filter vocabulary library to obtain a filter word; using a target filter corresponding to the filter word as a filter used after the live broadcast room starts broadcasting; the step of matching the word array with the preset filter vocabulary library to obtain the filter word comprises the following steps: extracting a target word with a target part-of-speech from the word array, wherein the target part-of-speech comprises adjectives and nouns; acquiring original text corresponding to the target word; 2. The method of claim 1, wherein, if it is determined that the target word is available by using the original text, matching the target word with the preset filter vocabulary library to obtain a filter word corresponding to the target word. the step of generating the word array matched with the live broadcast room according to the live broadcast room information comprises the following steps: dividing the live broadcast room information into words to obtain a plurality of first words; 3. The method of claim 2, wherein, respectively performing part-of-speech tagging on each first word to generate the word array matched with the live broadcast room. The live broadcast room information at least comprises introduction information and a room name of the live broadcast room, and the step of dividing the live broadcast room information into words to obtain a plurality of first words comprises the following step:

4. The method according to claim 2 or 3, characterized in that, dividing the introduction information and / or the room name into words to obtain a plurality of first words. The step of respectively performing part-of-speech tagging on each first word to generate the word array matched with the live broadcast room comprises the following steps: determining the part-of-speech of each first word; forming a first key-value pair by combining each first word and the part-of-speech corresponding to the first word; acquiring a TF-IDF word frequency feature and a tagging sequence feature of each first word, and calculating a weight coefficient of each first word by using the word frequency feature and the tagging sequence feature; 5. The method of claim 1, wherein, sorting the first key-value pair by using the weight coefficient to generate the word array matched with the live broadcast room. The step of extracting a target word with a target part-of-speech from the word array comprises the following steps: traversing the part-of-speech corresponding to the first word in the word array; 6. The method of claim 1, wherein, taking a first word with an adjective or noun part-of-speech as a target word. The step of matching the target word with the preset filter vocabulary library to obtain a filter word corresponding to the target word if it is determined that the target word is available by using the original text comprises the following steps: if there are an odd number of negative words corresponding to the target word in the original text, determining that the target word is unavailable; 7. The method of claim 1, wherein, if there are an even number of negative words corresponding to the target word in the original text, determining that the target word is available, and matching the target word with the preset filter vocabulary library to obtain a filter word corresponding to the target word. The step of matching the target word with the preset filter vocabulary library to obtain a filter word corresponding to the target word comprises the following steps: search, from a preset filter vocabulary library, a filter vocabulary same as, synonymous with, or similar to the target word as a filter word.

8. The method of claim 1, wherein, Further comprising: if there is no filter vocabulary matching the target word in the filter vocabulary library, a preset filter is selected as the filter used after the live room goes on air.

9. The method of claim 1, wherein, The filter vocabulary library is generated by: obtaining filters of the live room and at least one second word corresponding to the filters; forming a second key-value pair of the filters and the second words corresponding to the filters; generating the filter vocabulary library by using each second key-value pair.

10. The method of claim 9, wherein, The generating the filter vocabulary library by using each second key-value pair comprises: obtaining the frequency of use of each filter; sorting the second key-value pairs according to the frequency of use to generate the filter vocabulary library.

11. A live room filter selection apparatus, comprising: Comprise: a live room information obtaining module for obtaining live room information of a live room, the live room information comprising introduction information and room name, the introduction information being information added according to live content before going on air; a word array generating module for generating a word array matching the live room according to the live room information, the word array being generated by part-of-speech tagging of first words obtained by tokenizing the live room information, the part-of-speech of the first words comprising verbs, adjectives, nouns, and prepositions; a filter word obtaining module for matching the word array with a preset filter vocabulary library to obtain filter words; a filter determining module for determining a target filter corresponding to the filter words as the filter used after the live room goes on air; The filter word obtaining module comprises: a target word extracting submodule for extracting target words of target parts of speech in the word array, the target parts of speech comprising adjectives and nouns; an original text obtaining submodule for obtaining original text corresponding to the target words; a filter word obtaining submodule for matching the target words with a preset filter vocabulary library to obtain filter words corresponding to the target words if the target words are determined to be available by using the original text.

12. An electronic device, comprising: comprise a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete communication with each other through the communication bus; the memory is used to store computer programs; the processor is used to execute the programs stored on the memory to implement the method of any one of claims 1-10.

13. One or more computer-readable media having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method of any one of claims 1-10.

Images