12 | | As all Wiki pages, this page is editable, this means that you can |
13 | | modify the contents of this page simply by using your |
14 | | web-browser. Simply click on the "Edit this page" link at the bottom |
15 | | of the page. WikiFormatting will give you a detailed description of |
16 | | available Wiki formatting commands. |
| 7 | Project web: https://nlp.fi.muni.cz/projects/officebot/ [[BR]] |
| 8 | Project repository: [source: /nlp/projekty/officebot/trac-git] |
18 | | "[wiki:TracAdmin trac-admin] ''yourenvdir'' initenv" created |
19 | | a new Trac environment, containing a default set of wiki pages and some sample |
20 | | data. This newly created environment also contains |
21 | | [wiki:TracGuide documentation] to help you get started with your project. |
22 | | |
23 | | You can use [wiki:TracAdmin trac-admin] to configure |
24 | | [http://trac.edgewall.org/ Trac] to better fit your project, especially in |
25 | | regard to ''components'', ''versions'' and ''milestones''. |
| 10 | tasks: |
| 11 | * detect channel activity |
| 12 | * active/non-active (displayed as bold/normal typeface) |
| 13 | * type of activity (conversation = several mutually related posts, individual posts) |
| 14 | * language of the activity (Czech/English, short messages/long posts ...) |
| 15 | * sentiment of the activity (possible flamewar, emoji-based classification?) |
| 16 | * summarize activity for a particular channel |
| 17 | * from a particular moment track all messages |
| 18 | * cluster conversations (based on time span and/or topic) |
| 19 | * detect keywords |
| 20 | * highlight most important messages or part of messages |
| 21 | * summarize activity for all channels |
| 22 | * suggest joining a conversation based on user preferences (favorite topics, languages, sentiment ...) |
| 23 | * suggest moving part of a conversation into external applications |
| 24 | * transfer to calendar, trello, private channel (talk to myself) |
40 | | For a complete list of local wiki pages, see TitleIndex. |
| 42 | study Slack functionalities |
| 43 | |
| 44 | bring ideas about what can be done |
| 45 | }}} |
| 46 | {{{#!td |
| 47 | resolve legal issues concerning KM conversations data |
| 48 | }}} |
| 49 | |---------------- |
| 50 | || M2 || \ |
| 51 | {{{#!td |
| 52 | create the basic chatbot: |
| 53 | * listen on different channels |
| 54 | * detect channel activity |
| 55 | * quantify channel activity |
| 56 | }}} |
| 57 | {{{#!td |
| 58 | run the chatbot on KMLE Slack |
| 59 | |
| 60 | start collecting feedback |
| 61 | }}} |
| 62 | |---------------- |
| 63 | || M3 || \ |
| 64 | {{{#!td |
| 65 | study NLP techniques/implementations that can support smart features for the chatbot |
| 66 | |
| 67 | extend the chatbot with language detection |
| 68 | |
| 69 | propose a evaluation for the new feature |
| 70 | |
| 71 | cluster conversations based on time span |
| 72 | |
| 73 | }}} |
| 74 | {{{#!td |
| 75 | }}} |
| 76 | |---------------- |
| 77 | || M4-M5 || \ |
| 78 | {{{#!td |
| 79 | extend the chatbot with keyword detection using TF-IDF or another "simple" technique |
| 80 | |
| 81 | study the specifics of the Slack conversations |
| 82 | |
| 83 | cluster conversations based on keywords |
| 84 | |
| 85 | }}} |
| 86 | {{{#!td |
| 87 | }}} |
| 88 | |---------------- |
| 89 | || M5-M6 || \ |
| 90 | {{{#!td |
| 91 | study techniques for text summarization |
| 92 | |
| 93 | extend the chatbot with summarization |
| 94 | |
| 95 | }}} |
| 96 | {{{#!td |
| 97 | provide feedback for keyword detection |
| 98 | }}} |
| 99 | |---------------- |
| 100 | || M7 || \ |
| 101 | {{{#!td |
| 102 | extend the chatbot with sentiment analysis based on words/emojis |
| 103 | |
| 104 | suggest joining a conversation based on |
| 105 | keywords/topics/participants/sentiment... |
| 106 | }}} |
| 107 | {{{#!td |
| 108 | provide feedback for text summarization |
| 109 | }}} |
| 110 | |---------------- |
| 111 | || M4-M8 || \ |
| 112 | {{{#!td |
| 113 | extend the chatbot with connectors to applications such as Trello, calendar etc. |
| 114 | }}} |
| 115 | {{{#!td |
| 116 | provide feedback for sentiment analysis |
| 117 | }}} |
| 118 | |---------------- |
| 119 | || M9 || \ |
| 120 | {{{#!td |
| 121 | evaluate and deliver the application |
| 122 | }}} |
| 123 | {{{#!td |
| 124 | }}} |